Out-of-band management of rack mounted field replaceable units

Abstract

A system and method are provided that seamlessly provides centralized out-of-band (OOB) management of what may be many different field replaceable units (FRUs) within a rack-based system. A frame (100) is a virtual backplane for facilitating centralized OOB management of a plurality of FRUs (112), and includes frame arms (116) that are a plurality of virtual slots interconnected by a fixed topology. Each of the virtual slots detects when one of a plurality of FRUs is electrically interconnected to the virtual slot, and makes available to a frame center (120), a management controller electrically interconnected by the fixed topology, information indicating the virtual slot's specific location within the virtual backplane for use in performing OOB management of the FRUs.

Description

[Technical field]

[0001] FIELD OF THEINVENTION The present invention relates generally to managing field replaceable units (FRUs) mounted within a frame structure such as a rack or cabinet, and more particularly to a system and method for providing improved availability and serviceability to a rack-based computing system from a blade or chassis-based computing system. [Background technology]

[0002] Background technology "Out of band" (OOB) management (e.g., FRU discovery) for both rack-mounted servers (RMS) and blade compute systems Presence detection, inventory audit, power cycle and CPU reset activities such as discovery There have been great advances in recent years in the use of the Intelligent Platform Management Interface (IPMI) for system management, such as system IPMI is an industry standard computer system technology that provides an architecture or protocol that facilitates communication between a system controller or manager (e.g., including management software) and one or more unique devices (e.g., one or more FRUs) that are being managed.

[0003] A computing cabinet or rack is a standardized frame designed to hold multiple FRUs or related components (e.g., rack-mounted servers, power distribution units, or backup devices, etc.). Typically, a computing rack includes multiple (e.g., two, four) vertical rails or columns to which horizontal members and rail assemblies can be secured to define multiple receiving bays for receiving FRUs. FRUs of various types and sizes may be installed within the rack system and often have standardized heights as multiples of one rack unit (U). For example, industry standard rack systems are often 18U, 22U, 36U, 42U, etc. high. In high availability environments (e.g., telecommunications systems), a set of framed FRUs (e.g., computing devices, related components, etc.) are administered as a single compute system that is functionally compatible with the administration of a single FRU.

[0004] More recently, FRUs such as blade servers are used that are typically installed in a compartment or structure called a "blade enclosure" or chassis (e.g., in this case, the blade servers and the enclosure are collectively referred to as a "blade system"). The blade enclosure contains a midplane to which all of the blade servers are interconnected and provides a number of non-core computing services to the installed blade servers, such as power, cooling, interconnect, and management. That is, the installed blade servers collectively share such non-core computing services provided by the blade enclosure. For example, the blade enclosure may have a system controller or manager that includes any suitable control software or logic that functions to intelligently adjust a liquid cooling system to meet the cooling requirements of the blade servers. The system controller also facilitates the ability to "hot swap" blades within the enclosure (i.e., the ability to add, remove, and replace units as needed without having to power down the enclosure). The Advanced Telecommunications Computing Architecture (ATCA) is a family of modular, modular, and modular systems that allow for the creation of a single, highly flexible, and highly efficient system that can be used to integrate a single blade server into a single system. re) is an open industry standard that includes a set of specifications targeted at the requirements of blades and chassis (e.g., in terms of form factor, etc.). Summary of the Invention [Means for solving the problem]

[0005] Disclosure of the Invention Blade systems advantageously provide near 100% uptime or availability through the use of unified management services, redundancy or availability (e.g., if a first blade server goes down, a second blade server can take over and maintain the system without interruption to other blade servers until the first blade server is replaced), rapid fault isolation, low mean time to repair, midplane management paths to components, fixed FRU locations and configurations, reduced cabling, power and size requirements, etc. However, the blade enclosures and installed blade servers are typically proprietary designs. That is, a particular blade enclosure is usually designed to only accept a particular type of blade server, so that all of the blade servers installed within the blade enclosure have the same form factor, connectors, etc. Thus, the advantages and benefits of blade systems discussed above are inherently limited to a common type of FRU installed within a unique blade enclosure. Additionally, blade systems typically have a fixed power budget, cooling capacity, etc. Although future improvements to the blade design may be possible, any such future blades will be limited by the original blade design chassis for power and cooling.

[0006] The inventors have determined that it would be desirable to provide a rack-based system with many of the benefits of blade or chassis-based computing systems (such as the improved availability and serviceability discussed above), but without all or many of the limitations of blade-based systems. In other words, the inventors have determined that it would be desirable to provide the ability to seamlessly provide centralized OOB management (e.g., with respect to restart, shutdown, power on, fan speed, power and cooling monitoring, hot swap, etc.) of what may be many different FRUs (e.g., having many different form factors, power and cooling requirements, etc.) within a rack-based system. That is, it would be desirable to combine the high levels of serviceability and availability of blade-based systems with the improved capabilities of rack-mounted systems.

[0007] In this regard, disclosed herein is a system and method for managing FRUs mounted in a rack that provides the improved availability and serviceability of FRUs provided by blade-based systems in a manner that accommodates different types of FRUs (related to form factors, functionality, power and cooling requirements, etc.) installed in the rack or cabinet (e.g., server or open compute equipment racks, cabinets or assemblies for holding a plurality of computing devices, etc.). Generally, the disclosed system (e.g., a frame) includes a plurality of "frame backplane segments," "virtual slots," or "frame arms" (e.g., nodes, receiving structures, enclosures, connectors, etc., used interchangeably herein) disposable at fixed locations within the rack, each for receiving and electrically interconnecting a corresponding one of a plurality of FRUs (e.g., similar or different form factors, functionality, etc.) and each storing (e.g., in the virtual slot's memory) data indicative of a particular fixed location of the virtual slot within the rack (e.g., relative to the fixed locations of other virtual slots within the rack). Each virtual slot may be fixed to the rack adjacent a rear portion of a respective one of the plurality of receiving bays and may include a connector (e.g., a blind-mate connector) configured to interface with a corresponding connector on the FRU when the FRU is slid into the receiving bay. For example, each virtual slot may include , may be configured to detect when a FRU is inserted into its respective receiving bay (i.e., detect the presence of the FRU in the receiving bay) and send one or more corresponding alerts throughout the system as discussed below.

[0008] The disclosed system may include a "frame center" (e.g., a separate dedicated computing device or system, such as a centralized management server, which may also be a FRU) electrically interconnectable to each of the virtual slots in a manner that enables FRU presence (i.e., receipt of a FRU in one of the receiving bays) detection, FRU OOB management (e.g., via a "frame manager" or "rack manager" implemented in the frame center or otherwise in communication with the frame center), and the like. For example, the frame center may include a plurality of first type communication paths (e.g., I 2 C cable) to each of the virtual slots; The plurality of first type communication paths may form a first communication network interconnecting the frame center and the frame arms. The frame center may be electrically interconnected to each of the virtual slots by a plurality of second type communication paths (e.g., network lines such as Ethernet cables) that enable OOB management communications between the frame center and the frame arms, as well as non-OOB management communications between the frame arms and devices and processes external to the frame / system (through the frame center). Each of the virtual slots and the frame center may be electrically interconnected (e.g., via power lines or cables) to one or more (e.g., first and second redundant) power distribution units (PDUs, which may be suitably electrically interconnected to one or more power sources) for distributing power to FRUs that interface with the virtual slots (in addition to the frame center).

[0009] The various cables, paths, and / or lines connected between the virtual slots, the frame center, and the PDUs may be considered a "fixed interconnection topology" (e.g., wired and / or wireless) of the system or frame that enables the frame center to determine the physical or fixed location of installed FRUs in the rack (e.g., even among multiple different FRUs) via location information stored in the memory of their respective virtual slots for use in performing OOB management of the FRUs. A computing rack or cabinet may be pre-configured (e.g., pre-wired) with a fixed interconnection topology, such that a rack or cabinet that is subsequently (i.e., after pre-configuring) installed in the rack and (e.g., FRUs and frame art) may be configured with a fixed interconnection topology. FRUs that are electrically interfaced to their respective frame arms (through corresponding blind mating connectors in the rack) can substantially seamlessly or automatically receive power, participate in the rack's management network (e.g., as coordinated by the frame center), and / or send and receive actual data signals, all without having to necessarily route (e.g., manually) and interconnect multiple cords and cables between the FRUs and network switches, other servers, etc. (e.g., after and / or during the insertion of the FRUs into the rack). In this regard, the fixed interconnection topology and frame arms can essentially function as a "virtual backplane" or midplane for the system. The system may be built into a rack as part of the build of the rack, or may be retrofitted or otherwise integrated into an existing rack. The stem and rack are sometimes collectively referred to as the "setup."

[0010] As part of performing OOB management, the frame manager communicates with the OOB service processors (e.g., Oracle's Integrated Lights-Out Managers (ILMS)) of each of the FRUs through the frame center (e.g., the service processors in the frame center) and a fixed interconnect topology. The system may communicate with a frame manager (e.g., a frame manager for embedded system management tools, each including one or more dedicated processors and memory modules), such as a frame manager for embedded systems management tools (e.g., a frame manager for embedded systems management tools, each including one or more dedicated processors and memory modules). The system may incorporate a unified management system (UMS), consisting of various pieces of logic or software executed by the frame center, various FRU OOB service processors, and / or higher level processors. For example, by placing the FRU in a receiving bay of a rack and interconnecting the FRU to its respective frame arm, a portion of the UMS may be automatically downloaded to the FRU for use by the FRU's OOB service processor as part of communicating with the frame manager. In this regard, the FRU may be able to participate in its appropriate management service context (e.g., a UMS) substantially automatically and seamlessly, regardless of the type of FRU, the form factor of the FRU, etc.

[0011] As an example, assume that an FRU is installed in a receiving bay of a computing rack, thereby interconnecting the above-mentioned connector of the FRU to a corresponding connector of a particular frame arm of the receiving bay. Upon detecting the presence of the FRU (e.g., by circuitry in the frame arm, such as by detecting a power draw by the FRU), the frame arm notifies the frame center of the presence detection on a fixed interconnection topology (e.g., by each I / O interconnected between the frame arm and the frame center). 2 C The frame center (e.g., its service processor) may then allocate the memory of the virtual slot (e.g., 2 C line The frame manager may read the frame center address (e.g., in memory at the frame center) to obtain an indication of the fixed location of the virtual slot (and thus the installed FRU) within the rack, and then utilize the fixed location data to perform one or more OOB management tasks for the newly installed FRU. For example, the frame manager may maintain, or at least have access to (e.g., in memory at the frame center) various types of information related to the installed FRU's minimum processing and storage capacity, power and cooling requirements, physical location, current firmware version, network connections, etc., all of which may be used as part of the FRU's OOB management.

[0012] In this regard, part of the OOB management performed by the frame manager may include obtaining one or more properties from the OOB service processor of the newly installed FRU and comparing such properties to minimum or expected properties. Once the frame manager determines that the FRU contains the expected properties or is otherwise validating the FRU, the frame manager may then instruct the main processor on the FRU's motherboard to power up (e.g., via establishing a connection with the FRU's OOB service processor (e.g., over a respective Ethernet cable interconnected between the frame center and the frame arm)), which may allow the FRU to proceed to operate in any appropriate manner. The UMS portion on the FRU's OOB service processor may then send an alert or message to the frame manager over an appropriate communication path (e.g., a network cable) related to a fault, a FRU hot swap request, etc. Among other advantages, a rack or cabinet including the disclosed frames or systems installed therein may preserve data center floor space for subsequent rapid deployment of different types of FRUs (e.g., in response to service demands, introduction of new services, etc.).

[0013] In one aspect, a virtual backplane for facilitating centralized OOB management of a plurality of hot-swappable FRUs includes a plurality of virtual slots interconnected by a fixed topology, each of the virtual slots configured to: a) detect when one of a plurality of FRUs is electrically interconnected to the virtual slot; and b) store a virtual slot status information for use by a management controller, also electrically interconnected to the fixed topology, in performing OOB management of the FRUs. It includes circuitry that makes available information indicative of a particular location within the virtual backplane.

[0014] In one arrangement, the circuitry for each of the virtual slots includes memory that stores specific location information (e.g., address, bits, etc.) of the virtual slot relative to other virtual slots in the backplane. In another arrangement, the circuitry includes a number of indicators (e.g., LEDs) operable to indicate one or more statuses of the virtual slot or a FRU interfacing to the virtual slot (e.g., a FRU interfacing to the virtual slot, no FRU interfacing to the virtual slot, a FRU interfacing to the slot and active, an appropriate transition of the FRU from an initial state to a final state, one or more error conditions, etc.). In a further arrangement, each virtual slot of the virtual backplane includes a connector electrically interconnected to the circuitry, the connector adapted to interface to a corresponding connector on the FRU to facilitate communication between the FRU and both the virtual slot and the management controller.

[0015] In another aspect, a method of operating a virtual backplane to manage a plurality of FRUs in a computing rack includes detecting, in a first virtual slot of a plurality of virtual slots of the virtual backplane, the presence of an FRU electrically interfacing to the first virtual slot, sending an alert of the detected presence from circuitry of the first virtual slot to a centralized management module of the virtual backplane, and providing, by the circuitry of the first virtual slot, an indication of a location of the first virtual slot within the virtual backplane for use by the centralized management module of the virtual backplane.

[0016] In one arrangement, the method may include performing, by the centralized management module, one or more management routines associated with the FRU using the retrieved location indication. As an example, the centralized management module may utilize the retrieved location information as a lookup or key into a table or database of minimum computing requirements of the virtual backplane before bringing the FRU online or otherwise allowing the FRU to participate in the virtual backplane. For example, the centralized management module may communicate with the FRU (e.g., with the FRU's OOB service processor) over a LAN via the first virtual slot to determine whether the FRU meets any necessary minimum computing requirements of the virtual backplane.

[0017] In another arrangement, a first virtual slot may receive serial data communication from a FRU and then transmit the serial data communication to an I / O bus for use by the circuitry of the first virtual slot. 2 Convert to C protocol In one embodiment, the serial data communication may be a request to read from the first virtual slot's memory an indication of the first virtual slot's location in the virtual backplane (e.g., so that the FRU's OOB service processor knows relative to other FRUs and virtual slots where it is in the virtual backplane). In another embodiment, the serial data communication may be a request to read from the first virtual slot's memory an indication of the FRU's role in the virtual backplane (e.g., so that the FRU's OOB service processor knows relative to other FRUs what role it is supposed to play in the virtual backplane).

[0018] In a further arrangement, the first virtual slot may receive a request to remove the FRU from the first virtual slot (e.g., via a user pressing a button on the virtual slot, etc.) and then send an indication of the removal request to the centralized management module for processing. Upon receiving the removal request, for example, the centralized management module may proceed to perform any necessary OOB management tasks or routines to prepare for the actual removal of the FRU from the virtual backplane of the computing rack. In one embodiment, the centralized management module and / or any suitable indicator (e.g., LED, ) may be illuminated or otherwise activated to indicate that the FRU may be safely removed.

[0019] In a further aspect, a virtual slot of a virtual backplane is disclosed that facilitates OOB management of one of a plurality of field replaceable units (FRUs) mounted in one of a plurality of receiving bays in a computing rack by a centralized management controller connected to the computing rack. The virtual slot includes a memory that stores data indicating a location of the virtual slot relative to other virtual slots in the virtual backplane, and an I / O interface that provides access to the location data. 2 C data bus, serial data and I 2The virtual slots include a circuit board including logic to translate between the virtual slots' C data and the virtual slots' C data. The virtual slots also include connectors (e.g., blind mating connectors) electrically interconnected to the circuit board that are configured to interface with corresponding connectors on the FRUs to enable OOB management of the FRUs by the centralized management controller.

[0020] In one aspect, a system for delivering direct current (DC) power to a field replaceable unit (FRU) of a computing rack includes a rectifier having a plurality of input nodes for receiving a respective plurality of mains power sources and at least one output node for outputting a DC power source, and an electrical bypass mechanism. The electrical bypass mechanism includes an input node electrically interconnected to the output node of the rectifier, a first output node, a second output node, a first conductive path electrically connecting the input node and the first output node, and a second conductive path electrically interconnectable between a) the first input node or the first conductive path of the electrical bypass mechanism and b) a second output node. The first and second output nodes are configured to output respective DC voltages regardless of whether the system receives one or more mains power sources and / or one or more DC input power sources.

[0021] In one arrangement, the system may include a first power distribution unit (PDU) electrically interconnected to a first output node and a second PDU electrically interconnected to a second output node, where each of the first and second PDUs is configured to transmit DC power to a plurality of FRUs received within a computing rack. For example, a method for providing DC power to FRUs mounted within a computing rack may include electrically interconnecting the first PDU to a first port of each of the FRUs and electrically interconnecting the second PDU to a second port of each of the FRUs. A plurality of utility power sources may be electrically interconnected to a plurality of input nodes of each of the rectifiers. In one variation, the method may further include interrupting the flow of current between the first conductive path and the second conductive path (e.g., via removing a conductor from the second conductive path of the bypass arrangement or operating a switch disposed along the second conductive path) and electrically interconnecting a DC input power source to the second conductive path of the bypass arrangement or to the second output node. In another variation, the method may include electrically interconnecting a first DC input power source of the bypass arrangement to one of the first conductive path or the first output node and electrically interconnecting a second DC input power source to the second conductive path of the bypass arrangement or to one of the second output nodes.

[0022] In another aspect, a setup includes a cabinet having a plurality of receiving bays disposed within a framework of the cabinet, first and second DC PDUs secured to the framework, each configured to distribute DC power to FRUs mounted within the plurality of receiving bays, and an apparatus electrically connected to the first and second PDUs. The apparatus receives one or both of an AC input power source and a DC input power source and outputs a DC voltage to each of the first and second DC PDUs. In one arrangement, the apparatus includes a rectifier for receiving and rectifying the AC input power source and a bypass mechanism for receiving the rectified AC power source and / or the DC input power source, the bypass mechanism bypassing the rectifier to the first and / or second PDUs. In another arrangement, each receiving bay includes a respective electrical connector secured to a framework adjacent a rear portion of the receiving bay, each of the first and second PDUs being electrically connected to a respective one of the plurality of electrical connectors, and each electrical connector of the plurality of receiving bays is configured to interface with a corresponding electrical connector of a respective FRU mounted within the receiving bay to electrically interconnect the FRU to the first and second PDUs.

[0023] In a further aspect, a method for transmitting a DC voltage to a plurality of FRUs of a computing rack using an AC and / or DC input power source includes electrically connecting each of the plurality of FRUs to a respective first and second DC PDU, electrically connecting each of the first and second DC PDUs to an output of a power conversion device, and performing at least one of electrically connecting the AC input power source to an input of the power conversion device or electrically connecting the DC input power source to an output of the power conversion device. In response to performing, the power conversion device transmits the DC voltage to each of the first and second DC PDUs. In one arrangement, the performing step includes only electrically connecting the AC input power source to an input of the power conversion device, and the method further includes rectifying the AC input power source to a DC voltage and transmitting the DC voltage to the first and second DC PDUs via the output of the power conversion device. In another arrangement, the performing includes only electrically connecting a DC input power source to an output of the power converter, and electrically connecting the DC input power source to the output of the power converter includes electrically connecting a first DC input power source to a first output node of the output of the power converter and electrically connecting a second DC input power source to a second output node of the output of the power converter.

[0024] In one aspect, a fixed interconnect topology for implementing a backplane of a blade chassis in a computing rack includes a plurality of first type communication cables, each extending from a common first location to a respective one of a plurality of fixed locations and forming a first communication network of the fixed interconnect topology, a plurality of second type communication cables, each extending from the common first location to a respective one of the plurality of fixed locations and forming a second communication network of the fixed interconnect topology, and a plurality of power cables, each extending from a common second location to a respective one of the plurality of fixed locations.

[0025] In one embodiment, the common first location may be a centralized server, and each of the plurality of first ends of each of the plurality of first and second type communication cables may be interconnected to a respective one of the plurality of first and second type interfaces of the centralized server. For example, the plurality of first type interfaces may be I 2 C data port Alternatively, the plurality of second type interfaces may be Ethernet ports. As another example, each of a plurality of opposing second ends of each of a plurality of first type communication cables may be interconnected to a respective one of a plurality of memories in a plurality of fixed locations, each of the plurality of memories storing data indicative of a respective fixed location within the fixed interconnection topology, and a centralized management server utilizing one of the first type communication cables to read the fixed location data from a respective one of the plurality of memories of at least a first FRU of the plurality of electrically interconnected FRUs that is to perform at least one type of management.

[0026] In another aspect, a method for pre-wiring a computing rack to enable centralized management of multiple FRUs subsequently received within multiple receiving bays of the computing rack includes the steps of: 2 Connect the first end of each of the C cables to the centralized computing Electrically interconnecting devices and multiple I 2 Connect the second end of each of the C cables to The method includes electrically interconnecting each of the plurality of circuit boards to a respective one of the plurality of circuit boards and affixing each of the plurality of circuit boards adjacent to a respective one of the plurality of receiving bays of the computing rack. 2 One of the C cables, and obtaining data indicative of a physical location of each one of the receiving bays within the computing rack via one of a plurality of circuit boards secured adjacent to the receiving bay for use in performing one or more management routines in association with the FRU secured within each one of the receiving bays.

[0027] In one arrangement, the method includes electrically interconnecting a first end of each of the plurality of network cables to a centralized computing device, electrically interconnecting a second end of each of the plurality of network cables to a respective one of a plurality of connectors, and electrically interconnecting each of the plurality of connectors to a respective one of a plurality of circuit boards. In one variation, the method may further include electrically interconnecting a first end of each of the plurality of power cables to a power distribution unit, and electrically interconnecting a second end of each of the plurality of power cables to a respective one of the connectors.

[0028] In a further aspect, a virtual backplane implemented in a computing rack includes a plurality of I / Os, each having opposing first and second ends. 2 Includes multiple I / C cables 2 A first end of each of the I / O cables is electrically connected to a centralized computing device, 2 The second ends of each of the C cables are connected to a plurality of receiving bays of the computing rack. The virtual backplane also includes a plurality of power cables each having opposing first and second ends, the first ends of each of the plurality of power cables being electrically connected to the power distribution unit and the second ends of each of the power cables being electrically connected to one of a plurality of connectors each electrically connected to one of the plurality of circuit boards.

[0029] In one arrangement, the virtual backplane also includes a plurality of network cables each having opposing first and second ends, where the first ends of each of the plurality of network cables are electrically connected to the network switch and the second ends of each of the network cables are electrically connected to one of a plurality of connectors, each of which is electrically connected to one of the plurality of circuit boards. For example, each of the plurality of network cables may be electrically interconnected to a pass-through circuit of a respective one of the plurality of circuit boards.

[0030] In one aspect, a method for use in managing a plurality of FRUs mounted in receiving bays of a computing rack includes, in a management controller, a plurality of I / Os connected between the management controller and each of the circuit boards in the receiving bays of the computing rack. 2 The first FRU asserts I on one of the C data lines. 2 The method includes receiving a signal that one of the I / O data lines is to be inserted into or removed from one of a plurality of receiving bays to which the I / O data lines are connected, by a managing controller in response to the received signal. 2 On one of the C data lines, 2 The method also includes reading a memory of a respective one of the plurality of circuit boards connected to one of the C data lines to determine geographic location data of a first FRU within the computing rack, and performing, by the management controller, one or more OOB management routines for the first FRU using the geographic location data.

[0031] In one arrangement, the performing step includes retrieving, from a memory associated with the managing controller, a management record corresponding to the geographic location data, and managing, by the managing controller, the first FRU using the retrieved management record. For example, the managing may include receiving, at the managing controller, a first property of the first FRU from the first FRU, evaluating, by the managing controller, the retrieved management record associated with the received first property, and performing, by the managing controller, at least one action based on a result of the evaluating. As another example, the managing may include, by the managing controller, receiving a first property of the first FRU from the first FRU, a first property of the first FRU from the managing controller ... and performing, by the managing controller, at least one action based on a result of the evaluating. The information read may also include communicating with an OOB service processor of the first FRU over a network data line connected between the OOB service processor and a circuit board in one of the FRU bays to perform one or more OOB management routines for the first FRU using the read information.

[0032] In another arrangement, the method may include controlling, by the managing controller, one or more indicators resident adjacent one of a plurality of receiving bays of the computing rack or the managing controller based on the action taken. In a further arrangement, the received signal may be an interrupt generated by one of a plurality of circuit boards associated with the insertion or removal of the first FRU into or from one of the plurality of receiving bays.

[0033] In another aspect, a method for use with a plurality of FRUs mounted in a plurality of receiving bays of a computing rack includes reading a memory of a management controller of the computing rack to obtain a physical topology of a plurality of communication lines predicted to be electrically interconnected between the management controller and each of a plurality of receiving structures respectively secured adjacent to each of the plurality of receiving bays, each of the plurality of receiving structures adapted to interface with one of the plurality of FRUs. The method also includes determining, by the management controller, whether each of the plurality of communication lines is electrically interconnected between the management controller and each of the plurality of receiving structures according to the read physical topology.

[0034] In one arrangement, the determining may include sending a plurality of signals over the plurality of communication lines from the managing controller and verifying receipt of each of the plurality of signals by one of the plurality of receiving structures respectively associated with one of the plurality of communication lines in the retrieved physical topology. In another arrangement, the method may further include determining that at least one of the plurality of communication lines is not electrically interconnected between the managing controller and the plurality of receiving structures according to the retrieved physical topology and generating a signal based on the determining.

[0035] In a further aspect, an OOB service manager module that enables centralized OOB management of rack-mounted equipment in a computing rack includes a service processor module and an I / O module that is interconnected to the service processor module and includes a plurality of general-purpose input / output (GPIO) pins. 2 and a C data bus, each of the multiple GPIO pins having multiple I 2 and transmitting, on at least one of the C data lines, to or from one of a plurality of circuit boards disposed at fixed locations adjacent to a respective one of a plurality of receiving bays of the computing rack. 2 C signal and the OOB service manager module further includes a plurality of network ports interconnected to the service processor, each of the plurality of network ports configured to transmit and receive data to and from a plurality of rack-mounted equipment mounted respectively within the plurality of receiving bays via respective pass-through circuits of a plurality of circuit boards, and the OOB service manager module further includes a memory that stores a physical topology definition of the connections between the OOB service manager module and the plurality of circuit boards, the physical topology definition including data indicative of a fixed location of each of the circuit boards within the computing rack, and the service processor module utilizes the physical topology definition to provide centralized OOB management of the plurality of rack-mounted equipment mounted respectively within the plurality of receiving bays of the computing rack.

[0036] In one aspect, a method for electrically interconnecting one or more FRUs to at least one power source and at least one communications network in a computing rack includes receiving the FRU in a receiving bay of the computing rack, the FRU including a connector arrangement secured adjacent a rear portion of the FRU, the connector arrangement electrically interconnected to at least one power port and at least one communications port of the FRU. The method also includes, during at least a portion of the receiving, interfacing the FRU connector arrangement to an auxiliary connector arrangement secured adjacent a rear portion of the receiving bay of the computing rack, the auxiliary connector arrangement electrically interconnected to at least one power source and at least one communication network within the computing rack, and the interfacing electrically interconnects the FRU to the at least one power source and at least one communication network within the computing rack.

[0037] In one arrangement, each of the FRU connector arrangement and the auxiliary connector arrangement includes a blind mating connector and a mechanical connector spaced apart from the blind mating connector. The interfacing may then include contacting the blind mating connector of the FRU connector arrangement with the blind mating connector of the auxiliary connector arrangement to electrically interconnect the FRU to at least one power source and at least one communication network in the computing rack, and contacting the mechanical connector of the FRU connector arrangement with the mechanical connector of the auxiliary connector arrangement to align the blind mating connector of the FRU connector arrangement with the blind mating connector of the auxiliary connector arrangement. In another arrangement, the computing rack includes a plurality of receiving bays arranged along at least one dimension of the computing rack, a rear portion of each of the plurality of receiving bays including a respective auxiliary connector arrangement configured to interface with a respective FRU connector arrangement. In a further arrangement, at least one cable is secured between the auxiliary connector arrangement and a centralized management controller of the computing rack, and the method further includes, during interfacing, detecting, by the centralized management controller, a presence of the FRU in the computing rack via the at least one cable.

[0038] In another aspect, a system for automatically interconnecting a FRU to an OOB management network of a computing rack is disclosed, the computing rack including a plurality of receiving bays for receiving a plurality of FRUs. The system includes a rail member configured to slide relative to an auxiliary rail member adjacent one of the plurality of receiving bays of the computing rack, the rail member including a plurality of mounting components for fixedly securing the rail member to side portions of the FRU, and an adapter secured to the rail member for enabling blind-mating electrical interconnection of the FRU to the OOB management network of the computing rack. The adapter includes a mounting bracket fixedly secured to the rail member and a mounting member extending from the mounting bracket, the mounting member configured to receive a first blind mating connector, the first blind mating connector configured to interface with an auxiliary second blind mating connector secured adjacent a rear side of one of the plurality of receiving bays when the rail member slides relative to the auxiliary rail member.

[0039] In a further aspect, an FRU adapted for blind mating electrical connection into a receiving bay of a computing rack includes a housing having opposing top and bottom, opposing first and second side portions, and opposing front and rear portions, and first and second slide rail members secured respectively to the first and second side portions of the housing, each of the first and second slide rail members configured to interface with a complementary slide rail member secured adjacent to a side portion of the receiving bay, the FRU further includes an adapter bracket having a mounting portion and a mounting portion, the mounting portion being secured for movement relative to the housing and the mounting portion being spaced from a rear portion of the housing, the FRU further includes a blind mating connector secured to the mounting portion of the adapter bracket, and a plurality of cables. A first cable of the plurality of cables is electrically interconnected between the blind mating connector and the power port in the enclosure, and a second cable of the plurality of cables is electrically interconnected between the blind mating connector and the communication port of the enclosure, and the blind mating connector of the FRU is configured to interface with an auxiliary blind mating connector secured adjacent to a rear of the receiving bay, while the first and second slide rail members are configured to interface with the auxiliary slide rails of the receiving bay. The FRU slides against the rail members to electrically interconnect the FRU to the management network of the computing rack.

[0040] Any of the embodiments, arrangements, etc. discussed herein may be used with any of the disclosed aspects (alone or in combination with other embodiments, arrangements, etc.). Merely introducing a feature in accordance with a commonly accepted practice based on precedent does not limit the corresponding feature to the singular. The absence of use of a phrase such as "at least one" does not limit the corresponding feature to the singular. Use of a phrase such as "at least generally," "at least partially," "substantially," etc. in connection with a particular feature encompasses that corresponding characteristic and insubstantial variations thereof. Furthermore, reference to a feature in connection with the phrase "in one embodiment" does not limit use of the feature to a single embodiment.

[0041] In addition to the illustrative aspects and embodiments described above, further aspects and embodiments will become apparent by reference to the drawings and by study of the following descriptions. [Brief description of the drawings]

[0042] [Figure 1] FIG. 1 illustrates one embodiment of a setup including a smart computing frame or system installed in a computing rack to facilitate out-of-band management of multiple FRUs in the rack, such as servers, backup power modules, etc., mounted in the rack. [Diagram 2] 1 is a schematic diagram of a system according to an embodiment designed to accept AC and / or DC input power and distribute redundant DC voltages to each of a number of FRUs mounted within a computing rack. [Figure 3a] FIG. 3 is a schematic diagram of one configuration of the system of FIG. 2 designed to accept an AC input power source. [Figure 3b] 3 is a schematic diagram of another configuration of the system of FIG. 2 designed to accept a DC input power source. [Figure 3c] 3 is a schematic diagram of a further configuration of the system of FIG. 2 designed to accept AC and DC input power sources. [Figure 4] FIG. 2 is a more detailed schematic diagram of a frame arm of the frame of FIG. 1 intended to interface with a corresponding FRU; [Diagram 5] FIG. 2 is a more detailed schematic diagram of a frame center of the frame of FIG. 1. [Figure 6] 2 is a diagram of a representative sequence of events that may occur in the context of OOB management of a FRU in the frame of FIG. 1. [Figure 7] 2 is a flow diagram of a method for managing FRUs of a smart computing frame using worker and manager PROM images stored in the frame of FIG. 1. [Figure 8] FIG. 2 is a schematic diagram of a worker PROM image of the frame of FIG. 1. [Figure 9] FIG. 2 is a schematic diagram of a manager PROM image of the frame of FIG. 1. [Figure 10] FIG. 2 illustrates another embodiment of the smart computing frame of FIG. 1. [Figure 11-1] 2 illustrates another embodiment in which at least two of the smart computing frames of FIG. 1 are interconnected to form a smart computing frame system. [Figure 11-2] 2 illustrates another embodiment in which at least two of the smart computing frames of FIG. 1 are interconnected to form a smart computing frame system. [Figure 12] FIG. 2 is a rear perspective view of the example smart computing frame of FIG. 1. [Figure 13] FIG. 13 is a perspective view of a frame arm of the frame of FIG. 12. [Figure 14] FIG. 13 is a perspective view of an adapter that may be used to interface an FRU to a frame arm of the frame of FIG. [Figure 15] FIG. 15 is a perspective view of the adapter of FIG. 14 secured onto the side of an FRU. [Figure 16] FIG. 15 is another perspective view illustrating how the adapter of FIG. 14 facilitates electrical connections between various ports of the FRU and blind mating connectors. [Figure 17] FIG. 17 is an enlarged perspective view of the rear portion of the blind mating connector of FIG. 16. [Figure 18] 17 is a side view of the FRU of FIG. 16 received in a receiving bay of the frame of FIG. 12 before the blind mating connector of the adapter interfaces with a corresponding blind mating connector of the frame arm of FIG. 13. [Figure 19] 19 is a side view similar to FIG. 18, but after the blind mating connector of the adapter has interfaced with the corresponding blind mating connector of the frame arm of FIG. 13. [Figure 20]1 is a flow diagram of a method for electrically interfacing a FRU to a computing rack upon insertion of the FRU into a receiving bay of the computing rack. [Figure 21] FIG. 13 is a perspective view of an adapter and corresponding frame arm according to another embodiment. [Figure 22] FIG. 22 is another perspective view of the adapter and frame arm of FIG. 21. [Figure 23] FIG. 22 is another perspective view of the frame arm of FIG. 21. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0043] BEST MODE FOR CARRYING OUT THEINVENTION Disclosed herein are systems and methods for centralized management of FRUs (e.g., servers, backup power modules, etc.) mounted in multiple racks that provide similar levels of availability and serviceability as provided in chassis or blade-based systems, but without many of the inherent limitations of blade-based systems. Generally, the disclosed systems include multiple "virtual slots" or "frame arms" (e.g., receiving structures, enclosures, connectors, etc.) for electrically interconnecting corresponding multiple FRUs (e.g., of similar or different form factors and functionality). For example, adjacent frame arms may be appropriately spaced along the height or other dimension of a computing rack or cabinet (e.g., 1U or multiples thereof) and may be conveniently aligned with slide rails and / or other mounting features such that FRUs inserted into the rack via the slide rails or other mounting features may be operable to automatically interconnect with the respective frame arms.

[0044] The disclosed system may also include a "frame center" (e.g., "frame management module," "centralized management controller," "centralized management module," "centralized computing device," etc.) electrically interconnectable to each of the virtual slots by a "fixed interconnection topology" in a manner that allows FRU presence detection (i.e., detection of a FRU in one of the receiving bays at a known, fixed location in the rack), FRU OOB management (associated with startup, hot swapping, power and cooling monitoring, etc. via a "frame manager" implemented in and / or communicating with the frame center), etc. The system (including the frame arms, frame center, and fixed interconnection topology) may be incorporated into a rack or cabinet as part of building a rack, or may be retrofitted or otherwise integrated into an existing rack (i.e., an existing rack that is not specifically erected for a particular purpose for the system). That is, the rack / cabinet and FRUs may be standalone products independent of the system. The combination of the system / frame and the rack or cabinet may together form a "setup."

[0045] As used herein, the term "fixed interconnect topology" refers to a number of communication paths / lines / cables that fixedly connect the frame center to each of a number of frame arms. This implies multiple I / Os (e.g., those required for detecting FRU presence and fixed location, OOB management, etc.). Multiple communication lines mean multiple I / Os. 2 C cables, each of which are electrically interconnected between the frame center and a respective one of the frame arms. The multiple communication lines may also include multiple Ethernet cables, each of which is electrically interconnected between the frame center and a respective one of the frame arms. The fixed interconnection topology may also include multiple power paths / lines / cables that fixedly interconnect each of the one or more PDUs to each of the multiple frame arms and / or directly to each of the multiple installed FRUs. Each of the communication and power paths or lines includes two endpoints, with a first endpoint (e.g., endpoint "A") connected to a specific port of the frame center or PDU and a second endpoint (e.g., endpoint "B") connected to a specific frame arm (e.g., if the frame arm is at offset U in the frame).

[0046] 1, a schematic side view of a computing frame or system 100 is disclosed that enables rack-type FRUs (e.g., rack-mounted servers, backup power modules, etc.) to be managed in a manner similar to how blade-type servers are managed within a blade enclosure or chassis, but generally without many of the inherent limitations of blade-type systems. The frame 100 may be secured or otherwise implemented within any suitable rack or cabinet 104 having multiple bays or receptacles (not shown in FIG. 1) sized to receive, in a stacked or overlapping manner, multiple FRUs 112 of one or more types and / or form factors, such as rack-mounted servers, blade enclosures (each holding one or more blade servers), and / or other electronic devices. The rack 104 may include any suitable number of columns or vertical support members or rails 108 (e.g., two, four, etc.) based on the particular environment in which the frame 100 is to be implemented. The posts 108 may have associated therewith a series of sliding rails, rail assemblies, or other structures (not shown) that allow the FRUs 112 to be selectively inserted into and secured to the rack 104. The posts 108 may be suitably secured to a floor or other adjacent building structure to limit the frame 100 from tipping over. Other details regarding the rack 104 (e.g., panels, fasteners, etc.), how the FRUs 112 are inserted and removed from the rack 104, etc. will not be discussed further in connection with FIG. 1 for the sake of clarity and / or brevity.

[0047] Frame 100 includes a number of frame arms 116 (e.g., virtual slots, receiving structures, etc.), each of which is configured to electrically interface (e.g., via respective corresponding connectors 128, 132, as discussed in more detail below) with a corresponding one of a number of FRUs 112. Frame 100 also includes a frame center 120 (e.g., a separate dedicated computing device) that implements a frame manager 168 (e.g., software, logic) or otherwise acts toward the frame manager 168 for receiving FRU insertion / removal alerts from the frame arms 116 and performing OOB management of the FRUs 112 (e.g., if frame manager 168 is implemented on a separate device). While the frame manager 168 is illustrated as being implemented within the frame center 120, some arrangements may have some or all of the frame manager 168 implemented on a device separate from the frame center 120 but in communication with the frame center 120 (e.g., an installed FRU 112, a device separate from the frame center 120 and the FRU 112, etc.).

[0048] A fixed interconnection topology 124 (e.g., harness) of communication lines or paths (e.g., wired or wireless) directly interconnects the frame center 120 to each of the frame arms 116. As discussed above, multiple communication paths may be implemented using multiple I 2 C route (even Each of them may include a frame center 120 and a frame arm 1 16. Each I 2 C cables are The frame center 120 facilitates the transmission of alerts from each frame arm 116 to the frame center 120 regarding FRU presence detection, requests to remove a FRU 112 from a particular receiving bay (e.g., initiated via a user pressing a button on the respective frame arm 116), and the like. 2 Using the C cable, each frame The memory of frame arm 116 may also be read to obtain data indicating the fixed location of frame arm 116 (and therefore the interfaced FRU 112) within frame 100, and thereafter utilize the fixed location data as part of performing OOB management of the interfaced FRU 112 (discussed in more detail below).

[0049] The multiple communication paths of the fixed interconnection topology 124 may also include multiple network (e.g., Ethernet) paths (e.g., cables) each electrically interconnected between the frame center 120 and a respective one of the frame arms 116 and collectively creating a local area network (LAN) within the frame 100 and / or rack 104. Each network cable may essentially "pass" through its respective frame arm 116 to facilitate substantially direct communication between the frame center 120 and the OOB service processor 164 (e.g., ILOM) of the interfaced FRU 112 for use in performing OOB management of the FRU 112 (in addition to enabling network (e.g., Internet, WAN, LAN) communications between the FRU 112 and servers, processes, devices, etc. external to the frame 100 and / or rack 104). In this regard, the frame center 120 may be configured to provide an I / O interface for the FRU 112. 2 Using the C cable, you can read the memory of the frame arm 116 (for example). 1 through the network cable, and may receive FRU insertion / removal alerts and identify one or more specific fixed locations within the frame 100 of particular frame arms 116 and corresponding FRUs 112, and may perform OOB management of one or more of the FRUs 112 (e.g., via communication with the OOB service processor 164 and / or other management entities for the FRUs 112) utilizing the network cable.

[0050] For example, each cable / line may be in the form of "line endpoint (A): management server (X): port (N) <-> line endpoint (B): receiving structure (U)". The fixed interconnect topology 124 may also include a plurality of power paths / lines / cables that fixedly interconnect each of one or more PDUs 126 to each of a plurality of frame arms 116 (so that the FRUs 112 may receive power when the connectors 132 of the FRUs 112 interface with corresponding connectors 128 of their respective frame arms 116), or in some arrangements directly to ports on each of the FRUs 112. In any event, the fixed interconnect topology 124 may essentially form at least a portion of a "virtual" backplane or midplane into which the FRUs 112 can participate substantially seamlessly in whatever management service context it is part of (in this case the UMS running on the frame manager 168 and the OOB service processor 164 of the installed FRUs 112).

[0051] Before discussing frame arms 116, frame center 120, FRUs 112, and the interactions therebetween in more detail, reference is now made to Figure 2, which illustrates a system 500 for receiving one or both of an AC input power source (e.g., one or more utility power sources) and a DC input power source (e.g., a bank of batteries, an AC-DC converter, other DC power sources, etc.) and then providing first and second DC voltages (e.g., -48V) to respective first and second DC power distribution units (PDUs) for distributing dual / redundant DC power to FRUs mounted in a computing rack. Although system 500 will be discussed in conjunction with system 100, it should be understood that system 500 may also be used to provide power to FRUs mounted in a rack that does not incorporate system 100.

[0052] Generally, modern telecommunications systems are required to be able to accept either an AC power source (e.g., commercial electricity) or a DC power source (e.g., a bank of batteries) for powering the equipment of the system (e.g., FRUs mounted in a rack). In this regard, telecommunications systems often have different sets of equipment (e.g., power converters, rectifiers, PDUs, power cables, etc.) for respectively receiving AC and DC input power sources and distributing the voltage to the FRUs or other equipment of the system. For example, if a particular computing rack receives an AC power source, the rack has a first specific set of equipment designed to receive the AC power source and ultimately distribute (e.g., via one or more PDUs) the voltage to each of the multiple FRUs. However, if the same computing rack receives one or more DC power sources, the rack has a separate second specific set of equipment designed to receive one or more DC power sources and distribute (e.g., via one or more PDUs) the one or more DC power sources to each of the multiple FRUs. Maintaining separate sets of equipment for AC and DC power sources results in increased testing and validation costs for the manufacturer, increased operational costs for the manufacturer and end customer, and increased spare parts costs for both sets of equipment.

[0053] In this regard, the system 500 disclosed herein includes a single set of equipment designed to accept an input AC and / or DC power source and distribute at least one DC voltage to each of a number of FRUs mounted within a computing rack (e.g., FRU 112 of FIG. 1, FRUs of different computing racks, etc.), regardless of whether the input power source is AC and / or DC. Because the system 500 includes only a single set of equipment, it may realize numerous reductions in various types of costs (e.g., testing, operation, qualification, spare parts) associated with current battery communication systems. For example, because Network Equipment Building System (NEBS) qualification testing often costs in excess of $100,000 just for initial testing and up to $30,000 for subsequent testing, the cost savings of being able to utilize an AC and / or DC power source with a single set of equipment may be significant.

[0054] 2, the system generally includes an AC-DC converter or rectifier 504, at least first and second DC PDUs 508, 512 (e.g., rack mounted PDUs, smart or intelligent PDUs, etc.), and an electrical bypass mechanism 516 electrically connected between the rectifier 504 and the first and second DC PDUs 508, 512. The bypass mechanism 516 is configured to transfer a DC voltage to the first and second DC PDUs 508, 512 from an AC power source (e.g., via the rectifier 504) and / or from a DC power source as discussed below. In other words, the rectifier 504 and the bypass mechanism 516 may collectively be considered a "power conversion device" configured to transfer a DC voltage to each of the first and second DC PDUs 508, 512, regardless of whether the power source input to the power conversion device includes one or more AC power sources, one or more DC power sources, or both AC and DC power sources.

[0055] The rectifier 504 may generally include one or more (e.g., multiple) input nodes 520 (e.g., ports, contacts, pads) for electrical connection with one or more AC power sources or power supplies, such as one or more utility input feeds (e.g., 20 A, 200-240 V, single or polyphase, etc.), at least one output node 524 (e.g., port, contact, pad) for electrical connection to pass a DC voltage to the bypass mechanism 516, and circuitry 528 (e.g., any suitable arrangement of one or more transformers, diodes, resistors, etc.) configured to convert AC power from the AC power source to a DC voltage (e.g., −48 V). Each of the first and second DC PDUs 508, 512 may include at least one respective input node 532, 536 (e.g., , ports, contacts, etc.), a plurality of output nodes 540, 544 (e.g., outlets) for electrical connection to the frame arm 116 (e.g., to the connector 128 of the frame arm 116) or directly to the FRUs 112 or other equipment mounted in the rack, and any suitable circuitry 546, 550 operable to receive DC voltage from the bypass mechanism 516 and distribute DC power to the FRUs 112 (or other equipment mounted in the rack) via the output nodes 540, 544. Each set of power cables / cords 5541, 5542, 5543, 5544, etc. may be connected to a first and second DC Electrical connections may be made between respective output nodes 540, 544 of the PDUs 508, 512 and each of the frame arms 116, FRUs 112, etc.

[0056] For example, the first ends of the first set of power cables 5541 are connected to first and second D C. The first set of power cables 5541 may be plugged into or otherwise electrically connected to the first output nodes 540, 544 of the PDUs 508, 512. The second ends may both be electrically connected to the same frame arm 116, the same FRU 112, etc. Each of the first and second DC PDUs 508, 512 may be loaded to a percentage less than its rated maximum load to avoid tripping its circuit breaker. In one arrangement, the first and second DC PDUs 508, 512 may each share the load of the FRU 112 by approximately 50%, but each of the first and second DC PDUs 508, 512 is loaded to less than 50% of its rated maximum load. In this manner, if one of the first and second DC PDUs 508, 512 loses power, both of the first and second DC PDUs 508, 512 may be loaded to a percentage less than its rated maximum load to avoid tripping its circuit breaker. Even if the other of PDUs 508, 512 had to support 100% of the load, the load on the remaining PDU would still be less than its rated maximum load.

[0057] Continuing with reference to FIG. 2, the bypass mechanism 516 may include at least one input node 558 (e.g., contact, junction) for electrical connection (e.g., via trace, wire, cable) to the rectifier output node 524, at least first and second output nodes 562, 566 (e.g., contacts, pads) for electrical connection to the input nodes 532, 536 of the first and second DC PDUs 508, 512, respectively, a first conductive path 570 (e.g., trace, wire, cable) electrically connecting the input node 558 to the first output node 562, and a second conductive path 574 (e.g., trace, wire, cable) that is “interruptibly electrically connectable” between the input node 558 and the second output node 566. Although the second conductive path 574 is shown extending from the input node 558 to the second output node 566, it should be understood that in other embodiments the second conductive path 574 may extend from a portion of the first conductive path 570 (e.g., as just one example, from a midpoint of the first conductive path 570 between the input node 558 and the first output node 562) to the second output node 566 without departing from the spirit of the disclosure.

[0058] For purposes of this disclosure, "disruptibly electrically connectable" means capable of selectively interrupting or stopping the flow of current between the input node 558 and the second output node 566 for reasons discussed below. In one arrangement, the second conductive path 574 may be removable from the bypass mechanism 516. For example, the second conductive path 574 may be in the form of an electrical jumper that can be removed to interrupt the flow of current between the input node 558 and the second output node 566. In another arrangement, any suitable switch or button (e.g., not shown) may be disposed along the second conductive path 574 that is actuable or operable to selectively interrupt or not allow the flow of current between the input node 558 and the second output node 566. Other arrangements are possible and are within the scope of this disclosure.

[0059] Regardless of the input power source arrangement, the system 500 may be electrically interconnected to at least three different arrangements of input power sources and still provide a DC output voltage. Referring now to FIG. 3a (with the first and second DC PDUs 508, 512 omitted for clarity), one configuration of the system 500 is shown whereby the input power sources are a plurality of AC or utility input feeds 578 that are electrically connected respectively to a plurality of input nodes 520 of a rectifier 504. After the circuitry 528 rectifies the utility input feeds 578 to a DC voltage, the DC voltage is passed through an output node 524 of the rectifier 504 to an input node 558 of a bypass mechanism 516. The DC voltage is then sent along first and second conductive paths 570, 574 through first and second output nodes 562, 566 to the first and second DC PDUs 508, 512 for distribution to the FRUs of a computing rack or cabinet. Although one end of the second conductive path 574 is shown connected to the input node 558 , other arrangements are contemplated in which the end may be connected to any portion of the first conductive path 570 .

[0060] 3b, an alternative configuration is shown where the input power source is first and second DC power sources 582, 586 (e.g., each providing −40V to −60V) electrically interconnected via first and second conductive lines 590, 594 (e.g., cables, cords, etc.) to first and second output nodes 562, 566, respectively, of the bypass mechanism 516 to provide DC voltages to the first and second DC PDUs 508, 512, respectively. In this configuration, there is no AC input power source and any current flow between the input node 558 and the second output node 566 is interrupted.

[0061] In one arrangement, the second conductive path 574 may be removed (e.g., in the case of a removable jumper) and the first and second conductive lines 590, 594 may be directly connected, respectively, to the first and second output nodes 562, 566 of the bypass mechanism 516 (e.g., as shown in FIG. 3b) in any suitable manner (e.g., by a plug, hardwired connection, etc.). In another arrangement, the second conductive path 574 may remain within the bypass mechanism 516, but a switch or the like (not shown) disposed on the second conductive path 574 may be flipped or operated to interrupt the flow of current along the second conductive path 574. In this arrangement, the first conductive line 590 may be connected to the first output node 562 or to the first conductive path 570 (or even to the input node 558), while the second conductive path 594 may be connected to the second output node 566 or to the second conductive path 574 anywhere between the switch and the second output node 566. When the first and / or second conductive lines 590, 594 are connected to the first and / or second conductive paths 570, 574, respectively, the respective junctions between the first conductive path and the lines 570, 590 and between the second conductive path and the lines 575, 594 may be considered as the first and second output nodes 562, 566, in some cases. Because there is no AC input power source in this configuration, the rectifier 504 may be omitted in some embodiments (e.g., in contexts where it is contemplated that no AC input power source will be utilized). In one variation, the first conductive path 570 may additionally be appropriately interrupted (e.g., via removing the first conductive path 574, operating a switch disposed along the first conductive path 574, etc.).

[0062] FIG. 3c illustrates another configuration of the system 500 representing an integrated uninterruptible power supply (UPS) input. In this configuration, the flow of current between the input node 558 and the second output node 566 is again interrupted (e.g., via removing the second conductive path 574, activating a button or switch on the second conductive path 574, etc.) as discussed above in connection with FIG. 10b. However, the input power source now includes both AC and DC input power sources. For example, one or more utility input feeds 578 may be electrically connected to the input node 520 of the rectifier 504, respectively, to provide DC voltage to the first DC PDU 508 via the output node 524, the input node 558, the first conductive path 570, and the first output node 562. Also, a DC power source 598 (e.g., first and second DC A power source (one of 582, 586) may be electrically connected to the second output node 566 (or to a portion of the second conductive path 574 between a switch (not shown) and the second output node 566) via a conductive line 599 (e.g., one of the first and second conductive lines 590, 594) to supply a DC voltage to the second DC PDU 512.

[0063] Although the rectifier 504, the bypass mechanism 516, and the first and second DC PDUs 508, 512 are shown in Figures 2-3c as separate components, one or more of the various components may be integrated into other of the components. For example, the rectifier 504 and the bypass mechanism 516 (e.g., a power converter) may both be implemented in a single integrated circuit and / or a common housing. As another example, one or more features of the bypass mechanism 516 or the rectifier 504 may be implemented in the first and / or second DC PDUs 508, 512. In this regard, the diagrams shown in Figures 2-3c are presented merely to illustrate various functionality of the system 500, rather than necessarily limiting the breadth of the system 500. 1 , each frame arm 116 generally connotes any suitable arrangement of one or more parts or components that can electrically interface to one or more different types of FRUs 112 to enable the FRUs 112 to draw power from the PDU 126 and the frame center 120 to provide OOB management of the FRUs 112. Generally, each frame arm 116 may include a memory 136 that stores data 141 (e.g., one or more IDs or addresses) that indicate a particular fixed or physical location of the frame arm 116 within the frame 100 relative to the other frame arms 116. Each frame arm 116 may also have a connector 128 configured to mate with a corresponding connector 132 of the respective FRU 112 upon insertion of the FRU 112 into a respective receiving bay of the rack 104 (associated with the frame arm 116) to enable the FRU 112 to draw power from the PDU 126 and the frame manager 168 to communicate with the OOB service processor 164 of the FRU 112 via the fixed interconnect topology 124 and the respective connectors 128, 132.

[0064] 4 presents a more detailed schematic diagram of one of the frame arms 116 that is to interface with a corresponding FRU 112. The frame arm 116 may include any suitable housing 118 to which a connector 128 and a circuit board, such as a printed circuit board (PCB) 122, may be secured. The housing 118 may be mounted to a framework of the rack 104, for example, adjacent one of the receiving bays of the rack 104 (e.g., adjacent a rear portion of the rack 104), such that when the FRU 112 is inserted into the receiving bay, a connector 132 (e.g., a blind mating connector) of the FRU 112 may electrically interface (e.g., corresponding pins / contacts of the connectors 128, 132 may contact or otherwise electrically interface). The PCB 122 may have any suitable arrangement of circuitry including a memory 136 that stores location data 141 for the frame arms 116 (for example, the location data may be contained within a "worker" programmable read only memory (PROM) image 140, as discussed later in this discussion).

[0065] I 2 C data line 20 (e.g., a cable that is part of a fixed interconnect topology 124) 122 (e.g., an I / O cable, cord, trace, etc., not shown) is connected at one end to PCB 122 (e.g., 2 C data bus) at the opposite end (discussed in more detail below). The frame arm 116 is electrically connected to the frame center 120 to enable the frame arm 116 to send alerts to the frame center 120 (e.g., regarding FRU presence detection, a request to remove the FRU 112 from the receiving bay and therefore the OOB management network, etc.) and to enable the frame center 120 (e.g., upon receiving a corresponding alert) to read the location data 141 from the memory 136 and determine configuration information for the FRU 112 based on the read location information. PCB122 may be used to transmit serial data and I 2 Operable to convert between I / O and C data (e.g., I / O on PCB122) 2 C Data Bar The frame arm 116 may also include any suitable logic 130 (electrically connected to the frame center 120), one or more LEDs 180 (or other types of indicators) that generally indicate one or more operational states or statuses of the frame arm 116, and / or one or more buttons 182 that are operable, when manipulated (e.g., pressed), to cause the frame arm 116 to generate and transmit a signal or other communication to the frame center 120 (e.g., a request to remove the FRU 112 from the management network of the frame arm 116 and thus the frame 100). For example, the operational state of a particular frame arm 116 indicated by the LEDs 180 may range from an initial state in which no FRUs 112 are interfaced to the frame arm 116, all the way to a final state in which one or more FRUs 112 are interfaced and active (e.g., all manual servicing of the frame 100 is complete). The LEDs 180 may also indicate error conditions that may require an appropriate transition from the initial state to the final state and additional specific recovery actions.

[0066] The frame arm 116 supports a network (e.g., Ethernet) cable 24 from the frame center 120 to the connector 128, as well as first and second power cables 28, 32 (e.g., pair 5541, FIG. 2 ) from the first and second DC PDUs (e.g., first and second DC PDUs 508, 512, FIG. 2 ) to the connector 128. -5544) (one of the -5544)). As such, the network cable 24 and the first and second power cables 28, 32 do not necessarily need to communicate or interact with the PCB 122, and thus, when interfaced to the connectors 128, 132, may be substantially directly connected to the FRU 112.

[0067] With respect to the FRU 112 (a rear portion of the FRU 112 is shown in FIG. 4 ), the connector 132 may be fixedly (e.g., rigidly) secured to the FRU 112 via any suitable arrangement 134 (one representative example of the arrangement 134 is discussed below in connection with FIGS. 14-20 ) (e.g., directly or indirectly). The network (e.g., Ethernet) port 36 of the FRU 112 may be electrically connected to the connector 132 by a network cable 40, and the first and / or second power ports 44, 48 of the FRU 112 may be electrically connected to the connector 132 by respective power cables 52, 56. Additionally, the serial management port 60 of the FRU 112 may be electrically connected to the connector 132 by a serial data line 64. In this regard, interfacing to the respective pins and / or contacts (not shown) of connectors 128, 132 automatically enables FRU 112 to draw power from PDU 126 via power cables 28, 52 and / or power cables 32, 56, and automatically enables network communications between frame manager 168 (and / or frame center 120) and FRU 112's OOB service processor (e.g., ILOM) 164 (shown in FIG. 1) via network cables 24, 40 (as well as network communications between FRU 112 and devices / processes external to frame 100 and rack 104).

[0068] Additionally, upon interfacing to connectors 128, 132, a serial data connection is established between serial management port 10 and PCB 122 via serial data lines 64. At this point, a request from OOB service processor 164 of FRU 112 to read information (e.g., location data 141, role information, etc.) from memory 136 of PCB 122 (via serial management port 60 and serial data lines 64) is read by logic 130 from I / O 140 for use in servicing the request. 2 The requested information may be converted to C data. The data may then be converted back to serial data by logic 130 before being sent back to and / or received by OOB service processor 164 of FRU 112. It should be understood that each of the frame arms 116 may be substantially similar to the previously described frame arms 116, may differ only with respect to their fixed location data 141, etc. (i.e., each frame arm has a different specific fixed location within the frame 100), etc. However, each of the FRUs 112 that interface to the frame arms 116 need not necessarily be the same in terms of form factor, functionality, etc., so long as such FRU 112 includes a connector 132 matable with the connector 128 of one of the frame arms 116, and as discussed above, the connector 132 is electrically connected to the network port, serial management port, and power port of the FRU 112.

[0069] 5, a more detailed schematic diagram of the frame center 120 is presented. As discussed above, the frame center 120 may be a standalone, separate, dedicated computing device that facilitates OOB management by a frame manager 168 of what may be multiple different FRUs 112 mounted in a common computing rack 104 in a manner similar to how blades are managed in a blade chassis, but without many of the limitations inherently presented by blades (such as the need for all blades to have a common form factor, common functionality, etc.). In this regard, the frame center 120 generally includes a housing 121 that includes a management controller or service processor module 160 (including, e.g., a processor, on-board memory, etc., not shown) designed to work in conjunction with the frame manager 168 to provide OOB management of the FRUs 112. The frame center 120 also includes multiple interfaces 123 for electrical connection to each of multiple frame arms 116, and a PDU 126 through a fixed interconnection topology 124 for use in FRU insertion / removal detection, OOB management, etc.

[0070] For example, interface 123 may include multiple I / O pins, such as multiple general purpose input / output (GPIO) pins arranged on one or more bus expanders. 2 C-Interface 127 It may include 2 Each of the C interfaces 127 is Each I 2 C data line 20 to the PCB 122 of each frame arm 116 (See Figure 2.) 2 C interface 127 is 2 I C interface 127 electrically connected to service processor 160 2 C bus (Figure (not shown) or in any other way 2 Electrically interconnected to the C bus Interface 123 may also include a number of network interfaces 129, such as a number of Ethernet ports of a network switch that are electrically connected to service processor 160 in any suitable manner. Each network interface 129 is electrically interconnectable to a connector 128 of a respective frame arm 116 through a respective network line 24 of interconnect topology 124 (see FIG. 4 ).

[0071] Frame center 120 also includes (or at least has access to) a memory 144 that stores both fixed location data 141 of frame center 120 within frame 100 (i.e., the location of frame center 120 relative to frame arms 116 within frame 100) and fixed interconnect topology information 145. Generally, topology information 145 includes the fixed location data 141 of each of frame arms 116, the IP addresses of each of frame arms 116, the I / O ... 2 C-Interface 127 The topology information 145 includes a definition of the expected overall topology of the frame 100, including which of the network interfaces 129 are to be electrically connected to which of the frame arms 116 (e.g., each frame arm 116 may be identified by its respective fixed location data), which of the network interfaces 129 are to be electrically connected to which of the frame arms 116, etc. The fixed location data 141 and topology information 145 may also include configuration records for the FRUs 112 that are interfaced to the worker and manager PROM images 140, 148, respectively, as discussed later in this disclosure. For example, the configuration records may include information such as the types of FRUs 112 that may be interfaced to particular ones of the frame arms 116, the power and cooling requirements of the particular FRUs 112, etc. The fixed location data 141 and topology information 145 may be stored in the worker and manager PROM images 140, 148, respectively, as discussed later in this disclosure.

[0072] In one arrangement, the frame center 120 is 2 C interface 127 (for example) For example, 2 1. A data bus (not shown) electrically interconnected to one or more of the FRUs 112. The information 145 may also include a number of indicators 175, such as one or more LEDs 175 configured to activate (e.g., on, blinking, etc.) based on one or more conditions or statuses (e.g., fault, removal request, etc.). In any case, a user can flexibly customize the overall topology of the frame 100 (i.e., modify the "virtual backplane") and / or flexibly customize the fixed interconnect topology 124 (e.g., remove one end of a particular network line 24 from one of the network interfaces 127 and plug it into a different one of the network interfaces 127) simply by implementing changes to the information 145 (e.g., associating a particular frame arm 116 with a different one of the interfaces 123 of the frame center 120 via any suitable user interface that communicates with the frame center 120). The ability to flexibly customize the overall topology of a frame is in contrast to the backplane of a blade enclosure, which is fixed / static and generally cannot be modified.

[0073] In another arrangement, the service processor 160 of the frame center 120 may be able to verify whether the various connections between the frame center 120 and the frame arms 116 via the fixed interconnect topology 124 match an expected topology definition stored in topology information 145 in the memory 144 of the frame center 120. For example, the service processor 160 may 2 The first and second ends of the C wire 20 Each specific I specified in the topology information 145 2 C Interface 127 and 1, 4, and 5, the service processor 160 of the frame center 120 may first read the topology information 145 in the memory 144 to obtain a topology definition for the frame center 120, the frame arms 116, and the fixed interconnect topology 124. For example, the topology definition may include a first I / O bus 116 for the frame center 120, a second I / O bus 116 for the frame center 120, and a third I / O bus 116 for the frame center 120. 2 C interface 127 is the first I 2 The first frame arm 116 (as identified by its respective ID or location information 141) is electrically connected to the second I of the frame center 120 by the C line 20. 2 C Inter Faith 127 is the second I 2 C wire 20 electrically connects to second frame arm 116 The topology definition may specify which network lines 24 are to be connected to each other, etc. The topology definition may specify similar information about the various network lines 24 and / or other power and communication channels of the fixed interconnection topology 124.

[0074] The service processor 160 then negotiates the multiple communication lines (e.g., various I / O lines) of the fixed interconnect topology 124 according to the retrieved topology. 2 C and network line 2 0, 24) is electrically interconnected between frame center 120 and one of the frame arms 116. In one embodiment, service processor 160 may send signals over the communication lines (e.g., using the IP addresses of the frame arms 116, etc.) and then verify in any suitable manner receipt of each of the signals by one of the frame arms 116 that is respectively associated with one of the communication lines in the retrieved physical topology. In other words, The service processor 160 is configured to receive the specific I / O that is to be electrically connected to the frame arm 116 “#3”.2 A signal may be sent on the C line 20, then the frame arm 116# A similar process may be carried out on various other I 2 C and N This may be done for each of the network lines 20 and 24.

[0075] As an example, assume that the frame arms 116 are arranged in a particular orientation within the rack 104. For example, FIG. 1 illustrates how the frame arms 116 may generally be arranged in a vertically stacked manner between the top and bottom of the rack 104 (e.g., each frame arm 116 is physically maintained in a particular order within the rack 104, such as as a function of distance from the top or bottom of the rack 104, respectively). In this regard, the service processor 160 may sequentially send each signal on each respective communication line that is expected to be electrically connected to a particular frame arm 116 in a particular order that matches the arrangement of the frame arms 116 in the rack 104. Sequential receipt of the signals in the particular order may then be confirmed to quickly verify the correct wiring of the frame 100.

[0076] For example, each sequential signal may be configured to activate an indicator (e.g., LED 180) on each respective frame arm 116. In this regard, proper wiring of the frame 100 may be verified by a user observing a "visual walk" of the LEDs 180 from the top to the bottom of the rack 104 (or vice versa). Any wiring errors between the actual electrical connections and the expected electrical connections may be identified through jumps or inaccuracies in the sequential walk or activation of the LEDs 180 of the frame arms 116. As another example, the service processor 160 may read the topology information 145 (of FIG. 5) to determine which LEDs are expected to be electrically connected to the frame arms 116 in a particular order (e.g., from the top to the bottom of the rack 104). 2 C line 20 specific order Judging and determining the specific order of I 2The location data in the memory of the frame arm 116 is input via the C line 20. Then, the location data 141 is read out. 2 C line 20 The frame center 120 may be configured to verify that the location data matches the location data in the topology information 145 stored in the frame center 120. In one arrangement, any suitable indicator 175 (e.g., LED 177) on the frame center 120 may be configured to light up or otherwise activate depending on whether the frame is correctly wired. In response to a wiring error, the frame can be assessed, rewired, and retested to verify correct wiring.

[0077] In this regard, frame 100 persistently stores sufficient information to allow connected / interfaced FRUs 112 and frame center 120 to agree on the physical location of FRUs 112 within the frame (e.g., with respect to offset U). Such persistently stored information may reside in frame arms 116, frame center 120, fixed interconnect topology 124, and / or other suitable locations. For example, topology information 145 in memory 144 of frame center 120 may include a fixed listing of all lines / routes of fixed interconnect topology 124 in the form "Line End Point (A): Management Server (X): Port (N) <-> Line End Point (B): Receiving Structure (U)."

[0078] In any event, the service processor 160 and / or frame manager 168 utilize the fixed interconnect topology 124 and frame arms 116 to facilitate OOB management of the FRUs 112 that interface to the frame arms 116 (e.g., regardless of the manufacturer of the FRUs 112, the form factor of the FRUs 112, etc.). That is, the frame 100 is configured such that the service processor 160 and / or frame manager 168 can easily manage the FRUs 112 by cabinet 1, cabinet 2, and cabinet 3 (e.g., with respect to product type, motherboard revision, processor configuration, memory configuration, PCI root node and leaf node configuration, etc.). 100. The frame arm 116 may provide for OOB management of what may be many different types of FRUs 112 installed in the frame 100 in a substantially seamless manner, but in a manner similar to the manner in which a system controller in a blade chassis manages individual blades. Knowing the physical or fixed location of the FRUs 112 in the frame 100 allows the service processor 160 and / or frame manager 168 to easily pass communications and requests to and between the FRUs 112. Administrators may easily replace or otherwise correct malfunctioning FRUs 112 to maintain high levels of uptime, redundant fault architecture, low mean time to repair, and the like. In one arrangement, the service processor 160 and / or frame manager 168 may be able to monitor power draws to determine the location of the frame arm 116 and corresponding FRUs 112. For example, when a connector 132 of a FRU 112 interfaces with a corresponding connector 128 of a particular frame arm 116, the service processor and / or frame manager 168 may be designed to detect the resulting draw of power by the FRU 112 and thereby determine the fixed or physical location of the FRU 112 within the frame 100.

[0079] 6, one representative sequence of events 600 will be discussed in the context of OOB management of FRUs 112 in a frame 100. At 604, a signal may be received at the frame center 120 from one of the frame arms 116 for a FRU 112 that is electrically interfaced to the frame arms 116. As an example, assume that the FRU 112 is installed in the frame 100 such that its connector 132 interconnects / interfaces with a corresponding connector 128 of the frame arm 116 (e.g., as in FIGS. 1 and 4 ). The interface between the connectors 128, 132 triggers the transmission of a signal from the frame arm 116 to the service processor 160 of the frame center 120. For example, appropriate circuitry on frame arm 116 (e.g., circuitry on PCB 122, not shown) electrically connected to connector 128 may detect a power draw by FRU 112 (e.g., via power ports 44, 48 and power lines 52, 56, 28, 32 of FIG. 4) and then transmit a power signal (e.g., on an interrupt line electrically connected between frame arm 116 and frame center 120, and on a respective I 2 C data line 20 on top, etc.) frame center 12 1. In one embodiment, the FRU 112 may generate and send an interrupt or alert to the frame center 120's service processor 160. As another example, assume that a user desires to remove the FRU 112 from a receiving bay of the rack 104 to perform maintenance on the FRU 112, replace the FRU 112 with another FRU 112 (e.g., hot swap the FRU 112), etc. For example, the user may press a button 182 (e.g., see FIG. 4 ) on the frame arm 116, causing the generation and transmission of a “removal request” signal / alert to the frame center 120 (e.g., by any suitable logic / circuitry in the PCB 122, not shown).

[0080] In response to the received signal, the service processor 160 of the frame center 120 may proceed to determine 608 the ID (e.g., address, code, etc.) of the frame arm 116 from which the signal was received in any suitable manner, the determined ID distinguishing the frame arm 116 from other frame arms 116 in the frame 100. For example, the ID may identify a fixed location of the frame arm 116 within the frame 100, may be a unique number that identifies the frame arm 116 relative to other frame arms 116, etc. In one arrangement, the service processor 160 may determine the frame arm ID using the ID of the particular interface 123 through which the signal was received as a key into a table of interface IDs and corresponding frame arm IDs in the stored topology information 145 (see FIG. 3 ), etc. In another arrangement, the particular line / cable (e.g., I 2 C data line 20) is the frame The ID of the group 116 is read by the service processor 160 for use in determining the ID of the group 116. The cable may be a smart or intelligent cable that includes a memory that stores any suitable ID (e.g., a serial number or other identification data) that may be used to identify the frame arm. For example, service processor 160 may determine the frame arm ID using the smart cable's ID as a key into, such as, a table of smart cable IDs and frame arm IDs in stored topology information 145.

[0081] Using the determined frame arm ID as a key, the service processor 160 may retrieve 612 any appropriate OOB management records corresponding to the frame arm 116 from the topology information 145 in the memory 144 of the frame center 120 and manage 614 the FRU 112 using the retrieved management records. For example, the service processor 160 and / or the frame manager 168 may use the retrieved records to establish a connection with the OOB service processor 164 of the FRU 112 over a corresponding network line 129 electrically connected between the frame center 120 and the frame arm 116, and determine whether the FRU 112 meets or will meet any particular OOB management requirements of the frame 100 (e.g., minimum processing requirements and / or storage capacity, any particular motherboard revision number, etc., some or all of which may be policy driven).

[0082] As shown in FIG. 6, managing 614 may include determining 616 one or more properties of FRU 112 (e.g., communicating with OOB service processor 164 of FRU 112 to obtain the processing speed, storage capacity, etc. of FRU 112), evaluating 620 the obtained management records related to the one or more properties of FRU 112 (e.g., determining whether the FRU meets any specified minimum processing speed, storage capacity, etc.), and taking action 624 based on evaluating 620.

[0083] For example, upon determining that FRU 112 does not meet one or more minimum requirements, service processor 160 and / or frame manager 168 may not allow FRU 112 to participate in the management service context of frame 100 (e.g., and therefore may not even allow FRU 112's main processor to power up, or may only allow FRU 112 to proceed normally, i.e., as if FRU 112 were not part of frame 100). However, upon operationalizing FRU 112 (e.g., determining that the FRU does not meet any necessary management requirements), service processor 160 and / or frame manager 168 may instruct (e.g., via OOB service processor 164) the main processor on FRU 112's motherboard to power up so that FRU 112 can proceed to operate as part of the management service context of frame 100 (discussed more fully below).

[0084] In addition to FRU insertion alerts, frame center 120 may also receive alerts or messages from frame arms 116 related to fault conditions, requests to remove FRUs 112, etc., and take appropriate action. For example, when service processor 160 of frame center 120 receives an alert or other message indicating a fault condition from OOB service processor 164 of FRU 112 (e.g., on the respective network lines 24 electrically connecting frame center 120 to the frame arms 116 to which FRU 112 interfaces), frame manager 168 may take any appropriate remedial action, such as attempting to correct the fault, sending an alert to an administrator indicating the location in frame 100 of the faulty FRU 112, taking FRU 112 offline, etc.

[0085] As another example, if a user desires to disconnect a FRU 112 from its respective frame arm 116 (e.g., as part of a hot-swap operation), the user may press a particular button 182 (see FIG. 4) on the frame arm 116 to initiate a hot-swap from the frame arm 116 to the frame center service processor 160, as discussed above. Upon receiving the hot swap request, the service processor 160 may proceed to determine whether hot swapping of the corresponding FRU 112 is allowed. For example, a table of hot swap policies (or other management policies) for the various frame arms 116 and / or corresponding FRUs 112 (e.g., according to frame arm ID, such as location data) may be maintained in the topology information 155 in the memory 114 of the frame center 120. The service processor 160 may then proceed to allow or not allow the requested hot swap operation based on whether hot swapping of the FRU 112 installed in the particular frame arm 116 is allowed. In one arrangement, one or more LEDs 180 (see, e.g., FIG. 2) of a particular color on the frame arm 116 may be illuminated based on whether the requested hot swap operation is allowed. Numerous other examples of out-of-band management by the service processor 160 and / or frame manager 168 are contemplated and encompassed within the scope of this disclosure.

[0086] In one arrangement, a framework of management record specifications stored within programmable read-only memory (PROM) images in each of the frame center 120 and frame arms 116 may be provided and accessed by the service processor 160 and / or frame manager 168 to determine the location of the frame arms 116 and corresponding FRUs 112, administer management policies corresponding to particular FRUs 112, etc. As discussed, the disclosed management record specifications may be embodied in a "manager" PROM image 148 stored in memory 144 at the frame center 120 (and which may be accessed by the service processor 160 and / or frame manager 168 as part of managing the installed FRUs 112), and in multiple "worker" PROM images 140 stored in memory 136 of multiple frame arms 116 and in memory 144 of the frame center 120, and which may be accessed by the service processor 160 and / or frame manager 168 to perform OOB management, and by the installed FRUs 112 to determine frame role, physical location, etc.

[0087] Generally, installing a FRU 112 in a frame arm 116 causes the OOB service processor 164 of the FRU 112 to access a record in the worker PROM image 140 of the frame arm 116 (e.g., location data 141 in FIG. 2) to first determine whether the FRU 112 is installed in the frame 100 relative to another type of enclosure. Assuming the OOB service processor 164 is able to read the location record in the worker PROM image 140, it may then obtain information that defines both its location within the frame 100 and its role within the frame (e.g., whether the FRU simply functions as a "worker" FRU, or plays some type of management role within the frame 100, etc.). Additionally, service processor 160 and / or frame manager 168 obtain information (e.g., topology information 145) from manager PROM image 148 to understand how to manage each of the various FRUs 112 (e.g., in connection with hot swapping, propagating firmware updates, etc.), in addition to ensuring that updates to manager PROM image 148 include records corresponding to newly installed FRUs 112. Because product and configuration changes generally only require modifications to manager PROM image 148 as opposed to multiple worker PROM images 140, the disclosed management record specifications can advantageously remain largely static across different frame configurations, reducing management overhead.

[0088] 4, the memory 136 of each frame arm 116 may be in the form of a PROM that stores at least one corresponding worker PROM image 140 containing fixed location data 141 for the respective frame arm 116. Each worker PROM image 140 may be used by FRU 112 to store the location data 141 of the FRU 112 during frame 100 (i.e., when another The frame center 120 memory 144 may include information records for determining whether the FRU 112 is installed in the frame 100 (e.g., via fixed location data 141), determining the location of the FRU 112 within the frame 100, determining the particular "role" the FRU 112 should play within the frame 100, enabling the FRU 112 to obtain one or more IP addresses for communicating with the service processor 160 and / or frame manager 168, etc. (discussed below). Additionally, the frame center 120 memory 144 may be in the form of a PROM that stores at least one manager PROM image 148 that includes topology information 145 (i.e., information necessary to define the frame 100, relating, for example, to frame arm locations, configurations, power and cooling requirements, etc.) and that can be used by the service processor 160 and / or frame manager 168 to provide OOB management of the FRU 112, route incoming and outgoing communications between the frame arms 116, etc. The memory 144 of the frame center 120 also stores a corresponding worker PROM image 140 that the frame center 120 can use to determine its location within the frame 100, to determine the IP address of the FRU for routing management communications, etc.

[0089] To further facilitate the reader's understanding of how the FRUs 112, frame arms 116, and frame center 120 interact within the frame 100 to provide the aforementioned levels of increased availability and serviceability, reference is now additionally made to FIG. 7, which illustrates a method 200 for use with the frame 100, and FIGS. 8-9, which illustrate schematic diagrams of worker and manager PROM images 140, 148 for use within the frame 100. The method 200 may include interfacing 204 the FRU 112 to the frame arms 116 of the frame 100. For example, interfacing 204 may entail inserting the first FRU 152 into a particular bay of the cabinet 104 such that the connector 132 of the first FRU 152 interfaces or otherwise interconnects with the connector 128 of the first frame arm 156 (see FIG. 1 ). The method 200 may then include attempting 208 (e.g., by the OOB service processor 164 of the first FRU 152) to read a location record (e.g., location data 141 of FIG. 4) of the worker PROM image 140 of the first frame arm 156.

[0090] 8, the worker PROM 140 may include a location record 304 and a subnet IP record 324. The location record 304 may include one or more unique location IDs that serve to identify the geographic address or location of the frame arm 116 where the worker PROM image 140 is stored, e.g., to identify the geographic address or location of the corresponding FRU 112. For example, the location record 304 may include a "physical" location ID 308 that identifies the location within the frame 100 of the FRU 112 that is directly or physically interconnected to the connector 128 of the corresponding frame arm 116 (e.g., a FRU 112 connected to frame arm 116#2 may have a corresponding physical location ID 308 of "2," while a FRU 112 connected to frame arm 116#6 may have a corresponding physical location ID 308 of "6"). In one arrangement, the physical location ID 308 may provide a particular offset U (e.g., a number of offset rack units) within the frame 100.

[0091] The location record 304 may also include one or more "non-physical" location IDs 312, each of which identifies the location of at least one FRU 112 that is indirectly interconnected to the corresponding frame arm 116 via another FRU 112 that is directly interconnected to the frame arm 116. For example, FIG. 10 illustrates another embodiment of a frame 100' in which a connector 132 of a FRU 112' in the form of a blade enclosure interfaces to a connector 128 of a corresponding frame arm 116. The FRU 112' includes multiple FRUs 112" in the form of blade servers suitably mounted within the FRU 112'. In this case, the location record 304 of the corresponding worker PROM image 140 of the frame arm 116 would identify the location of the FRUs 112 in the frame 100'. The OOB service processor 164 may include a physical location ID 308 identifying a location of FRU 112″, and a number of non-physical IDs 312 identifying locations within the frame 100′ of the number of FRUs 112″. For example, a chassis management module (CMM) (not shown) of an Oracle Sun Netra 6000 chassis (e.g., FRU 112″) connected to frame arm 116 #8 may be identified by a physical location ID 308 of “8,” and the OOB service processor 164 of a blade server (e.g., FRU 112″) disposed in “slot #3” of the chassis may be identified by a non-physical location ID 312 of “131” (0×83).

[0092] If the location record 304 of a particular worker PROM image 140 includes one or more non-physical location IDs 312, the location record 304 may also include a list of subordinate location IDs 316 that map or otherwise link the non-physical location IDs 312 (i.e., IDs that identify the FRU 112″ to the frame 100) to local device numbers 320 (i.e., IDs that identify the FRU 112″ to the FRU 112′) and that may be used by the FRU 112″ to configure the FRU 112″ in a manner that does not require waiting for communication with the service processor 160 and / or frame manager 168. The location record 304 may also include an “identification” byte 332 that identifies the role of the FRU 112 within the frame (e.g., role as a frame manager 168, role as a worker FRU, etc.). For example, the identification byte 332 of the worker PROM image 140 of the frame center 120 may be set to the “frame manager” role, so that when the frame center 120 is placed in the frame 100 , the frame center 120 proceeds to function as the frame manager 168 .

[0093] The subnet IP record 316 includes a base management subnet IP address 328 that generally provides the OOB service processor 164 of each FRU 112 with the information necessary to communicate with the frame center 120, the service processor 160, and / or the frame manager 168, and / or other FRUs 112. More specifically, individual addresses (e.g., of other FRUs 112) may be derived by using the subnet IP address 328 as the network address and the physical or non-physical location ID 308, 312 as the host address.

[0094] 1 and 7, the read attempt step 208 of the method 200 may include the OOB service processor 164 of the first FRU 152 attempting to read the physical and / or non-physical location IDs 308, 312 of the location record 304 of the worker PROM image 140. The method 200 may then determine 212 whether the location record 304 can be read. In response to a negative determination at 212, the method 200 may proceed to 216, at which point the first FRU 152 may operate in a “normal” mode (e.g., a mode of operation without association with a management service context of the frame 100). In response to a positive determination at 212, the method 200 may proceed to 220, at which point the first FRU 152 may operate in a “smart frame” mode (e.g., a mode of operation at least partially controlled or governed by a management service context of the frame 100). Part of the consequence of a positive determination at 212 may be that the frame manager 168 instructs the main processor on the first FRU 152 to power up.

[0095] 7, operation 220 of the first FRU 152 in smart frame mode may include determining 224 the role of the first FRU 152 in the frame 100, such as reading and interpreting a bit mask of an identification byte 332 of the worker PROM image 140 via the OOB service processor 164. For example, in response to a positive determination to a query 228 as to whether the role is a worker FRU, the method 200 may proceed to 232, where the first FRU 152 is generally subordinate to the frame manager 168 (e.g., in a manner in which the first FRU 152 is managed by the frame manager 168). In response to a negative determination to query 228, method 200 may proceed to query 236 as to whether the determined role is a frame manager.

[0096] More specifically, although frame manager 168 is generally described and shown as being implemented by frame center 120, this is not necessarily always the case. For example, a bit mask of identification byte 332 of first FRU 152 of a worker PROM image may indicate a frame manager role. Upon determining such role, method 200 may proceed to retrieve 244 a management record from manager PROM image 148 of first frame arm 156 and then manage 248 the worker FRU (and possible additional FRUs) of frame 100 in accordance with the retrieved management record (e.g., via frame manager 168 executing on first FRU 152 in association with service processor 160 of frame center 120).

[0097] For example, at initial configuration of the frame 100 and / or at any other suitable time, an administrator or other user may load or otherwise store a manager PROM image in the memory 136 of the selected frame arm 116 (in addition to setting a bit mask of the identification byte 332 of the worker PROM image 140 of the selected frame arm 116 to correspond to the frame manager role). In one arrangement, each FRU 112, upon determining that it has the role of a frame manager, may store a copy of a frame manager 168 in memory 158 which it may execute. In another arrangement, upon determining that an FRU 112, upon determining that it has the role of a frame manager, may obtain a copy of a frame manager 168 from any suitable location (e.g., another FRU 112, a frame center 120, a higher level component, etc.) via a communication channel 124 for storage in memory 158 and subsequent execution. FRUs may operate 240 according to various other types of roles encompassed within the scope of this disclosure.

[0098] In either case, the service processor 160 and / or frame manager 168 utilize the information in the manager PROM image 148 as part of managing the FRUs 112 of the frame 100. Referring now to FIG. 9, the manager PROM image 148 includes location records 304 of all of the worker PROM images 140 associated with the installed FRUs 112 and frame centers 120, as well as subnet IP records 324. In this regard, the service processor 160 and / or frame manager 168 can derive individual addresses of each of the installed FRUs 112 and frame centers 120 using the location records 304 and subnet IP records 316 for use in communicating therebetween. The manager PROM image 148 may also include configuration data records 340, power and cooling requirements records 348, frame arm address records 352, and the like.

[0099] The configuration data record 340 generally stores configuration information specific to the FRU 112 installed in a particular frame arm 116 in the frame 100. As shown, the configuration data record 340 may store the configuration information through a number of "policy" bytes 344 that are tagged (e.g., with a particular physical location ID 308) to a particular location record 304. For example, the configuration data record 340 may include a "hot swap" policy byte 344 with a particular bit mask for each location ID 308 that indicates whether hot swapping of the FRU 112 associated with the location ID 308 is disabled, enabled, deactivated, etc. As another example, the configuration data record 340 may store configuration information specific to the FRU 112 installed in a particular frame arm 116 in the frame 100. , may include a "firmware" policy byte 344 that specifies whether the firmware should be checked to determine if it is of a particular version, whether it should be automatically upgraded, whether it does not need to be checked, etc. Numerous other types of policy bytes 344 are contemplated, such as a "memory size" byte 344 (e.g., specifying the minimum required available memory in an installed FRU 112 of a particular frame arm 116 as identified by physical location ID 308), an "installation type" byte 344 (e.g., specifying one or more particular types of FRUs that may be installed in a particular frame arm 116 as identified by physical location ID 308), a "network connection" byte 344 (e.g., specifying one or more particular types of network connections that an installed FRU 112 must have), etc.

[0100] The configuration data records 340 (as well as other records disclosed herein) may be arranged and organized in any suitable manner. For example, the configuration data records 340 may be arranged in a table or database format, whereby the physical and / or non-physical location IDs 308, 312 may fill the first row or first column, each of the various policy bytes 344 may fill others of the first row or first column, and a bit mask of the various policy bytes 344 may fill a cell in the table for each of the location IDs. The service processor 160 and / or the frame manager 168 may access the configuration data records 340 as part of managing the FRUs 112 installed in the frame 100. In one arrangement, the configuration of the manager PROM image 148 may reflect the FRUs 112 currently installed in the frame 100. More specifically, if no FRU 112 is directly interfaced to the frame arm 116, then the manager PROM image 148 may also be devoid of information specific to the FRU 112 to be installed in the frame arm 116 and the worker PROM image 140 of the frame arm 116 (e.g., location record 308, policy byte 344, etc.). Also, when a FRU 112 is installed in a particular frame arm 116 to join a frame management service context or network, the service processor 160 and / or frame manager 168 may facilitate updating the manager PROM image 148 to reflect information specific to the FRU 112 and frame arm 116 that are installed in the frame arm 116. In another arrangement, the manager PROM image 148 may store information specific to the frame arm 116, policy byte 344, etc., whether or not the FRU 112 is installed on the frame arm 116.

[0101] While the information stored in the manager PROM image 148 with respect to different frame configurations may change (e.g., to accommodate new or different configurations), the information in the worker PROM image 140 may remain substantially static. For example, consider a first configuration in which half of the frame arms 116 of the frame 100 are populated with FRUs 112, with each of the populated frame arms 116 including a respective worker PROM image 140. Further consider a second configuration of the frame 100 in which one or more of the previously unpopulated frame arms 116 of the frame 100 are now populated with one or more FRUs 112. Here, the manager PROM image (stored in the frame center 120 and / or one of the FRUs 112) may be updated to accommodate the newly added FRU 112, while each of the worker PROM images 140 corresponding to the installed FRUs 112 common between the first and second configurations may remain the same. However, it should be understood that when a particular FRU 112 installed in a frame arm 116 is replaced with a different FRU, the information in the worker PROM image 140 (and thus the manager PROM image 148) in the frame arm 116 may, in some circumstances, change accordingly. For example, when a 2U FRU 112 installed in one or two frame arms 116 is replaced with a 4U FRU installed in the same one or two frame arms 116, the information in the worker PROM image 140 (and thus the manager PROM image 148) in the frame arm 116 may change accordingly. When replaced with FRU 112, the location record 304 of the worker PROM image 140 of the frame arm 116 may be appropriately updated to reflect the different "rack location height" (e.g., due to height differences between 2U FRUs and 4U FRUs).

[0102] As an example of how the service processor 160 and / or frame manager 168 manage the FRUs 112 in which they are installed, assume that a first FRU 152 runs the frame manager 168 and that a second FRU 172 (see FIG. 1 ) interfaces to a second frame arm 176 of the frame 100. Assume further that the OOB service processor 164 of the second FRU 172 is able to read the location record 304 of the worker PROM image 140 stored in the second frame arm 176 and determines (e.g., upon reading the identification byte 332) that it is a worker FRU. The frame manager 168 then reads the respective I / O address 302 (e.g., indicating the insertion of the second FRU 172) and the I / O address 304 (e.g., indicating the insertion of the second FRU 172) and the ... 2 via C-line 20 Upon determining that a worker FRU (i.e., the second FRU 172) has been installed in the frame (by receiving an alert from the service processor 160 over the respective network line 24 that the service processor 160 has received an interrupt signal from the second FRU 172), the frame manager 168 may access the manager PROM image 148 of the first frame arm 156 to obtain a management record tagged with the physical location ID 308 of the second frame arm 176 for use in managing the second FRU 172.

[0103] For example, when the frame manager 168 receives a request from the OOB service processor 164 of the second FRU 172 to join the network (i.e., the network of the frame 100), the frame manager 168 may query the OOB service processor 164 of the second FRU 172 for various specifications or properties of the second FRU 172, such as the current firmware version, available memory, network connection, motherboard revision, product type, PCI root node and leaf node configuration, etc. Upon receiving the specifications, the frame manager 168 may analyze or evaluate the received specifications relative to a particular policy byte bit mask associated with the second frame arm 176, physical location ID 308, and take one or more actions based on the results of the evaluation. As an example, when the frame manager 168 determines that the second FRU 172 needs to update its current firmware to a newer version, the frame manager 168 may not allow the second FRU 172 to participate in the network of the frame until the second FRU 172 updates its firmware. As another example, upon receiving a request from the OOB service processor 164 of the second FRU 172 to bring one or more applications online (e.g., available for use by other FRUs 112 and / or higher level components or processes), the frame manager 168 may query the OOB service processor 164 about the current network connections of the second FRU 172 (e.g., which switch the second FRU 172 is connected to) and may or may not allow the applications to be brought online based on the results of the network connectivity query by the frame manager 168.

[0104] In one arrangement, when a user desires to hot swap the second FRU 172, the user presses or otherwise operates a button or other feature (e.g., button 182 in FIG. 4) on the back of the second frame arm 176 to activate each of the I / Os of the fixed interconnect topology 124 that electrically connects the second frame arm 176 to the frame center 120.2 C line 20 1. The hot swap request may be transmitted to the service processor 160 of the frame center 120 via the network lines 24 between the frame center 120 and the first frame arm 156. Upon receiving the request, the service processor 160 may forward the request to the frame manager 168 in the first FRU 152 (or to another In response to the frame manager 168 determining whether a hot swap of the second FRU 172 is allowed or not, the frame manager 168 may access and evaluate a "hot swap" policy byte 344 in a configuration data record 340 of the manager PROM image 148 that is tagged with the physical location ID 308 associated with the second frame arm 176 to determine whether a hot swap of the second FRU 172 is allowed, and then correspondingly allow or disallow the hot swap based on the results of the evaluating. For example, an LED 180 on the second frame arm 176 may assume a respective first and second state (e.g., blinking or not blinking) in response to the frame manager 168 determining whether a hot swap of the second FRU 172 is allowed or not allowed.

[0105] As discussed above, the frame center 120 is interconnected to each of the frame arms 116 by a fixed interconnection topology 124 to generally facilitate routing of communications between the installed FRUs 112. In this regard, the frame center 120 (e.g., the service processor 160) accesses a frame arm address record 352 in the manager PROM image 148 to determine how to route a particular communication to a particular frame arm 116 (and thus a FRU 112 installed in a particular frame arm 116). In one arrangement, the frame arm address record 352 may map a physical hardware address and a port number to a physical location ID 308. For example, when a FRU 112 attempts to bring one or more applications online, the OOB service processor 164 of the FRU 112 may generate a request that includes a destination physical or non-physical location ID 308, 312 associated with the FRU 112 of the frame manager and pass the request to the frame center 120 (e.g., on the respective network lines 24 that electrically interconnect the frame center 120 to the frame arm 116 to which the FRU 112 interfaces). Upon receiving the request at the frame center 120, the service processor 160 may utilize the destination physical or non-physical location ID 308, 312 to obtain a corresponding physical hardware address and port number (e.g., a hardware path) and then route the request to such obtained physical hardware address and port number on the particular network line 24 that electrically connects the frame center 120 to the frame arm 116 associated with the obtained physical hardware address and port number. In one arrangement, the service processor 160 propagates changes to the subnet IP records 316 to the frame arms 116 for updating of the respective worker PROM images 146 .

[0106] 10-11 show other embodiments of the frame 100. As previously discussed, FIG. 10 shows an embodiment including a plurality of FRUs 112″ in the form of blade servers, in which the connectors 132 of the FRUs 112′ in the form of a blade enclosure or chassis interface with the connectors 128 of the corresponding frame arms 116, and the FRUs 112′ are suitably mounted within the FRUs 112′. FIG. 11 shows a smart frame system 396 in which first and second 397, 398 frames (e.g., frames 100) installed in first and second racks 104, respectively, may be interconnected by any suitable communication channel or interconnection 399 (e.g., cables, harnesses) for use in increasing processing and memory capacity, etc. This In this arrangement, a single frame manager 168 implemented in the frame center 120 or FRUs 112 of either the first or second frames 397, 398 may manage all of the FRUs 112 installed on the first and second frames 397, 398. That is, one of the first and second frames 397, 398 may not have a frame manager 168. For example, communications between a frame manager 168 residing on the first frame 397 and a FRU 112 on the second frame 398 may be passed to the frame center 120 of the second frame 398, which (in conjunction with the manager PROM image 148) analyzes the communications and determines which FRU 112 the communications are directed to. The process proceeds to determine whether the route should be routed.

[0107] In one arrangement, the manager PROM image 148 may conform to the IPMI Platform Management FRU Information Storage Definition Version 1.0. Provided below is an example of a management record specification for use with the frame 100 disclosed herein, with the understanding that the disclosure is not limited to the specific form presented below. Rather, it is provided only as an example to provide the reader with one manner in which the disclosure may be implemented.

[0108] Primary Record Location record

[0109] [Table 1]

[0110] There can be four types of "Place ID" in a place record.

[0111] [Table 2]

[0112] Each Location ID may be 2 bytes in length and may have the following bit definitions:

[0113] [Table 3]

[0114] The "Rack Location" column may be filled in with the corresponding physical location ID. The Frame ID may be utilized in multi-rack configurations. The "Identification Byte" of the Location Record may be used to determine role information and may be used by the Frame Center during the programming process. The Identification Byte may contain a bitmask with the following definitions:

[0115] [Table 4]

[0116] The frame manager bits do not have to be mutually exclusive: for example, a worker FRU that has frame manager functionality in one logical domain (LDOM) and worker functionality in another LDOM might have an identification byte value of 0x03, while a worker FRU without frame manager functionality might have a value of 0x02.

[0117] The dependent location ID list may contain a series of device number-location ID mappings. The list may terminate with a location ID of zero. This information may be used by an entity FRU (e.g., an enclosure or chassis) along with a dependent attachment FRU (e.g., a blade server) so that it can configure its attachment without having to wait for frame manager communication. Software may ensure that these mappings match the corresponding FRU configuration data record (shown below). If the frame manager location ID is non-zero in the location record, it holds the location ID of a non-physical frame manager (e.g., one that is not directly connected to a frame arm, such as a blade in a blade server). If there is no corresponding frame manager in this attachment, the entry has a value of zero.

[0118] [Table 5]

[0119] Subnet IP Record The base management subnet IP address may be stored in the worker PROM image of each frame arm to give each FRU's OOB service processor the information it needs to communicate with the frame manager. An individual management address may be derived by using the base management subnet IP as the network address and the location ID as the host address. The frame center software may ensure that changes to this address are propagated to each frame arm location and frame center image.

[0120] [Table 6]

[0121] Configuration Data Record The configuration data record may hold most of the information specific to each location ID. The frame manager may access this data and compare it to the FRUs that are installed. FRU components that are intended to be checked may implement a policy byte that the frame manager evaluates to determine what action to take if the component does not meet minimum requirements (e.g., not enough memory installed).

[0122] [Table 7]

[0123] Frame arm address record The frame arm address record maps a physical hardware address and port number to a frame arm identifier (e.g., a location ID), which may be referenced by a frame center. This record may contain a list of structures terminating in a frame arm identifier of zero.

[0124] [Table 8]

[0125] The address map list may contain a series of location ID-hardware address map associations. The list may terminate with a location ID of zero.

[0126] [Table 9]

[0127] Option Recording FRU I / O Port Definition Record The FRU I / O Port Definition records describe each I / O port for a given location record. These descriptions may be used in point-to-point connection records (discussed below) and may accommodate FRUs with many ports across multiple records. The I / O port definition record has the following details:

[0128] [Table 10]

[0129] An I / O port list may contain a series of port ID-port type associations. The port ID may be unique for a given location ID. The port list for a specific record number may terminate with a port ID of zero.

[0130] [Table 11]

[0131] Point-to-point connection records A point-to-point connection record defines all I / O port interconnections within a frame. Each connection entry may contain a policy byte that the frame manager can use to determine whether the connection should be verified and what action to take if the connection verification fails.

[0132] [Table 12]

[0133] The connection list may contain a series of connection records that have point-to-point associations. For a given record, the connection list may terminate with a connection ID of 0. The software may ensure that all point-to-point connection records are evaluated.

[0134] [Table 13]

[0135] Dynamic Management Data Dynamic management data may be used by the management framework and may be considered dynamic in that it may change more frequently than the configuration. Dynamic management data may be stored in frame manager configurable files rather than PROM images. may be held.

[0136] [Table 14]

[0137] Policy byte definition

[0138] [Table 15]

[0139] Type Identifiers and Port Type List

[0140] [Table 16]

[0141] 12, there is shown a rear portion of one representative example of a smart computing frame 400 (e.g., frame 100 of FIG. 1) that may be used to efficiently manage a wide range of FRUs, such as rack mounted servers, backup power modules, and the like, in one or more form factors, such as 1U, 2U, etc. As shown, the frame 400 may be installed or otherwise disposed within a rack or cabinet 404 that includes width, height, and depth dimensions 405, 407, 409, and may generally be comprised of a plurality of vertical support members or rails 408 interconnected by a plurality of horizontal support members 411 to form a framework. The rack 404 may include a plurality of pairs of rail assemblies 401 disposed along a height dimension 407 (or other dimension) of the rack 404, with each rail assembly 401 extending along a depth dimension 409 of the rack between opposing vertical members 408. Each pair of rail assemblies 401 may define one of a number of receiving bays 410 for receiving one of a number of FRUs 412 .

[0142] Adjacent a rear portion of each of the receiving bays 410, the frame 400 may include a respective frame arm 416 (e.g., frame arm 116) that facilitates electrical interconnection of each FRU 412 to a frame center (not shown, e.g., frame center 120) and one or more PDUs (not shown, e.g., PDUs 508, 512) through a fixed interconnect topology 424 (e.g., fixed interconnect topology 124). Referring now to FIG. 13, an enlarged perspective view of one of the frame arms 416 is provided. The frame arm 416 may be secured in any suitable manner to one outer rail member 450 of a pair of rail assemblies 401 of the receiving bay 410 adjacent the rear portion of the receiving bay 410. In this regard, FRUs 412 (not shown in FIG. 13) seated in receiving bay 410 via outer rail members 450 may be automatically aligned with frame arms 416 such that continued insertion of FRUs 412 eventually results in an interface between FRUs 412 and frame arms 416, as discussed below. Interfaces and electrical connections can result.

[0143] The frame arm 416 generally includes a housing 418 (e.g., housing 118) that can be fixedly secured in any suitable manner to one outer rail member 450 of a pair of rail assemblies 401 of a particular receiving bay 410. In one arrangement, the housing 418 can include a pair of grooves 419 adjacent its lateral portions that are configured to receive a corresponding pair of flanges 421 of the outer rail member 450. In this regard, the housing 418 can be slid along the outer rail member 450 to a desired location (e.g., adjacent a rear portion of the receiving bay 410) and then fixedly secured to the outer rail member 450 in any suitable manner (e.g., via tightening bolts (not shown) that extend through the housing 418 relative to the outer rail member 450). Other manners of securing the housing 418 to the rack 404 adjacent a rear portion of the receiving bay 410 are also contemplated and encompassed within the scope of the present disclosure.

[0144] The frame arm 416 includes a PCB 422 (e.g., PCB 122) that is affixed to a portion (e.g., side) of the housing 418 (e.g., via passing fasteners 417 through apertures (not shown) in the PCB 422 and into apertures (not shown) in the housing 418) having a memory that stores location data of the frame arm 416 within the frame 400; 2 C conversion logic, LEDs, etc. (not labeled, but similar to PCB122 in Figure 4) ) and a connector 428 (e.g., connector 128, such as a blind mating connector) electrically connected and secured to PCB 422 in any suitable manner to facilitate electrical connections between each FRU 412, the frame center, and one or more PDUs via fixed interconnection topology 424. More specifically, at least one I 2 C cable or wire (for example, I 2C line 20, not shown in FIG. 13) is between the frame center and PCB 422 (e.g., through the bottom portion of PCB 422 facing outer rail member 450). 2 C data bus, electrically connected (not visible in Figure 13) , at least one network cable or line (e.g., network line 24, not labeled in FIG. 13 ) may be electrically connected between the frame center and pins / contacts (not shown) adjacent the rear portion 470 of connector 428 (and thus may “pass through” PCB 422), and at least one power line or cable (e.g., power lines 28, 32) may be electrically connected between one or more PDUs (e.g., PDUs 508, 512) and the pins / contacts adjacent the rear portion 470 of connector 428.

[0145] To facilitate precision alignment (not shown in FIG. 13 but discussed below in connection with FIGS. 14-19 ) between the connector 428 of the frame arm 416 and the corresponding connector 432 of the FRU 412, the frame arm 416 may include at least one mechanical connector or alignment component in the form of an alignment pin 429 configured to be received in a corresponding mechanical connector or alignment component in the form of an alignment barrel 740 of an adapter 700 (e.g., a “frame backplane adapter”) of the FRU 416 (discussed below). The alignment pin 429 may be affixed to a portion of the housing 418 such that it faces in the same direction as the connector 428 (i.e., toward the direction in which the approaching FRU 416 is inserted into the receiving bay 410), but is at least slightly spaced from the connector 428. For example, the alignment pin 429 may include a locking or retaining nut 430 in addition to a threaded end (not shown) that may be inserted through at least one aperture 431 in a wall of the housing 418. Threaded nuts (not shown) may be threaded over the threaded ends of alignment pin 429 onto opposite sides of the wall of housing 418 such that nuts 430 abut or contact the housing wall to rigidly secure alignment pin 429 to frame arm 416. In one embodiment, the threaded nuts may be removed from the threaded ends and the threaded ends are inserted into different ones of the apertures 431 in the wall of housing 418. 429 to achieve the desired height of the alignment pin 429.

[0146] As shown, the alignment pin 429 may include a tip portion 437 (which may be tapered, for example) to facilitate initial positioning of the alignment pin 429 in the alignment barrel 740 of the adapter 700 of the FRU 412. The alignment pin 429 may also include an enlarged head 435 adapted to interact with a periphery of the alignment barrel 740 to at least partially lock or secure the alignment pin 429 against movement relative to the alignment barrel 740 upon interfacing with the frame arm 416 and the connectors 428, 432 of the FRU 412. Any suitable cover, shield, or the like (not shown) may be secured to the housing 418 over the PCB 422 (e.g., via fasteners 417) to protect various components of the PCB 422 from damage. In one arrangement, one or more support arms or brackets (not shown) may rigidly interconnect the housing 418 of the frame arm 416 to opposing outer rail members 450 of opposing rail assemblies 401 of the same receiving bay 410 to further support the frame arm 416 and provide a stronger frame 400 and / or rack 404.

[0147] 14-15, perspective views are presented of an adapter 700 (e.g., a connector arrangement) configured to allow insertion of a FRU 412 into a receiving bay of a computing rack and to be electronically discovered substantially seamlessly by the computing rack's management services software by way of a blind mating interface between the adapter 700 and the receiving bay (e.g., frame arm 416 of receiving bay 410) of the computing rack. For example, the adapter 700 may at least partially represent the arrangement 134 of brackets, links, etc. discussed in connection with FIG. 4 that is adapted to immovably secure the connector 132 to the FRU 112 such that the connector 132 can electrically interface with the connector 128 of the frame arm 116 upon insertion of the FRU 112 into the receiving bay of the computing rack 104. The adapter 700 may be added to an existing standard rack-mounted FRU to quickly adapt the FRU for use in a blind-bond installation (such as in frame 400 of FIG. 12), or may be built into the FRU as part of the initial build of the FRU so that the FRU can be used in such a blind-bond installation.

[0148] Generally, the adapter 700 includes a mounting portion 704 (e.g., one or more brackets) configured to be fixedly attached to the FRU 412 and a mounting portion 708 (e.g., one or more brackets) extending from the mounting portion 704 configured to receive a connector 432 (e.g., blind mating connector 132 of FIG. 4 ) as the FRU 412 is inserted into the receiving bay 410 to enable the connector to electrically interface with a corresponding connector 428 (e.g., connector 128) adjacent the rear of the receiving bay 410 of the computing rack. In one arrangement, the mounting portion 704 may include a first mounting member 712 (e.g., a plate, a flange, etc.) configured to be fixedly attached to the FRU 412. For example, the first mounting member 712 may be firmly secured (e.g., via fasteners, nuts, and aligned holes, not labeled) to an inner portion of the inner rail member 454 of the same rail assembly 401 to which one of the frame arms 416 of the frame 400 is secured (e.g., the frame arm 416 is secured to the outer rail member 450 of the rail assembly 401 and the first mounting member 712 is secured to the inner rail member 454 of the same rail assembly 401).

[0149] In this regard, the rigid mounting of inner rail member 454 to side portion 413 of FRU 412 (e.g., via threaded fasteners (not shown) through mounting components such as apertures 458 in inner rail member 454 and corresponding apertures (not shown) in side portion 413 of FRU 412) immovably or rigidly secures adapter 700 to FRU 412. In another arrangement, mounting portion 704 may additionally or alternatively include a second mounting member 716 (e.g., a plate, flange, etc.) spaced from inner rail member 454 and configured to be fixedly secured to side portion 413 of FRU 412 (e.g., via a threaded fastener (not shown) through aperture 720 in second mounting member 716 and a corresponding aperture (not shown) in side portion 413 of FRU 412).

[0150] 14-18, the mounting portion 708 may include a mounting member 728 (e.g., a plate, bracket, flange, etc.) extending from the mounting portion 704 and having a receiving aperture 732 extending therethrough (labeled in FIGS. 14 and 17) sized and configured to fixably receive the connector 432 therein such that the forward portion 438 of the connector 432 is directed toward the forward portion 472 of the connector 428 of the corresponding frame arm 416 upon insertion of the FRU 412 into the receiving bay 410. For example, the connector 432 may include one or more flexible protrusions, clips, etc. (not shown) adapted to snap past an inner wall of the receiving aperture 732, thereby locking or fixedly securing the connector 432 to the mounting portion 708 and thus the FRU 412 (i.e., when the mounting portion 704 is fixed to or at least fixedly secured relative to the FRU 412). However, other manners of securing connector 432 to mounting portion 708 are contemplated and encompassed herein.

[0151] In any event, it can be seen how the mounting portion 704 and inner rail member 454 space the mounting portion 708 of the adapter 700 from the rear portion 414 of the FRU 412 by a distance 724 (labeled in FIGS. 16 and 18 ) to advantageously provide space for one or more cables, wires, etc., that electrically connect the power, serial, and network ports adjacent the rear portion 414 of the FRU 412 to the connector 432 secured to the mounting portion 708 of the adapter 700. For example, FIGS. 16-18 show how a number of cables 736 (e.g., network line 40; power lines 52, 56; serial data line 64 of FIG. 4 ) may each electrically connect various ports 737 (e.g., network port 36; power ports 44, 48; serial management port 60) of the FRU 412 to appropriate pins / contacts (not shown) adjacent the rear portion 433 of the connector 432. As shown, the first and second mounting members 712 , 716 and the attachment member 718 may collectively form at least a partial housing for housing a plurality of cables 736 . In either case, the electrical interface of the respective pins / contacts adjacent the forward portions 434, 472 of the connectors 432, 428 (e.g., upon insertion of the FRU 412 into the receiving bay 410 of the rack 404 as discussed below) automatically enables the FRU 412 to draw power from one or more PDUs 126, enables network communications to occur between a frame manager (not shown; e.g., frame manager 168 of FIG. 1), a frame center service processor (not shown, e.g., service processor 160 of FIG. 5), and the FRU 412's OOB service processor (not shown; e.g., ILOM 164 shown in FIG. 1), and enables requests from the FRU 412's OOB service processor to read information (e.g., location data 141, role information, etc.) from the PCB 422's memory (not labeled).

[0152] As discussed above, the frame arm 412 may include at least one mechanical alignment component in the form of an alignment pin 429 configured to be received within a corresponding mechanical alignment component in the form of an alignment barrel 740 of the adapter 700 to facilitate precise and secure alignment and electrical interface between the connector 428 of the frame arm 416 and the connector 432 of the FRU 412. With reference to FIGS. 14, 16, and 17, the alignment barrel 740 may be configured to be received within a corresponding mechanical alignment component in the form of an alignment barrel 740 of the adapter 700 adjacent or spaced apart from the connector 432. 416 and 416a, and may be disposed in or on the mounting portion 708. As shown, the alignment barrel 740 may include opposing forward and rearward portions 744, 748 and a receiving opening or bore 752 extending through the alignment barrel 740 between the forward portion 744 and the rearward portion 748. Although the alignment pin 429 and the alignment barrel 740 have been discussed as being disposed on the frame arm 416 and the adapter 700, respectively, other arrangements are contemplated where the alignment pin 429 and the alignment barrel 740 may instead be disposed on the adapter 700 and the frame arm 416, respectively.

[0153] To facilitate the reader's understanding of how a FRU (e.g., FRU 412) may be inserted into a computing rack (e.g., rack 404) and permitted to participate in the rack's management network in a substantially seamless manner (e.g., as a blind mating), reference is now made to FIG. 20. FIG. 20 illustrates one method 800 of inserting a FRU into a computing rack and electrically interfacing the FRU to a frame center of the rack. However, it should be understood that other methods involving more, fewer, or alternative steps are also contemplated and encompassed herein. In addition, although method 800 is discussed in the context of FRU 412 being inserted into a receiving bay 410 of rack 404 and electrically and mechanically interfacing to frame arms 416 of receiving bay 410, it should be understood that method 800 may be practiced in association with FRUs, computing racks, etc. other than those specifically illustrated herein.

[0154] At 804, ends 462 (e.g., see FIG. 16 ) of a pair of inner rail members 454 of the FRU 412 may be inserted into channels 451 of a pair of outer rail members 450 of a receiving bay 410 of a computing rack 404, respectively (e.g., via a front portion of the receiving bay 410 that faces a frame arm 416 disposed adjacent a rear portion of the receiving bay 410). For example, a first one of the inner rail members 454 may be secured to both a first side portion 413 of the FRU 412 (as shown in FIGS. 14-17 ) and the adapter 700, and a second one of the inner rail members 454 may be secured to an opposing second side portion (not shown) of the FRU 412. In one arrangement, a pair of inner rail members 454 may each be inserted into a channel of a pair of intermediate rail members 456 (see FIG. 18 ) that are respectively disposed within a channel 451 of a pair of outer rail members 450 such that each set of inner, intermediate, and outer rail members 454, 456, 450 slide or otherwise move relative to one another to facilitate insertion and removal of FRUs 412 into and from the receiving bay 410. Although an end 462 of one of the inner rail members 454 is illustrated as extending past an end of the adapter 700 (e.g., past the ends of the mounting and attachment portions 704, 708, see FIGS. 15, 16, and 18 ), other arrangements encompassed herein contemplate that the end 462 extends just to or even short of the end of the adapter 700.

[0155] The method 800 may then include advancing (e.g., sliding, moving, etc.) 808 the FRU 412 toward a rear portion of the receiving bay 410 (e.g., pushing a front portion of the FRU 412 such that the opposing rear portion 414 of the FRU 412 moves toward a frame arm 416 adjacent the rear portion of the receiving bay 410). See FIG. 18 , which illustrates the rear portion 414 of the FRU 412 as it is advanced 808 toward the frame arm 416 adjacent the rear of the receiving bay 410. The method 800 may then include mechanically aligning 812 the electrical connector 432 of the FRU 412 with the corresponding electrical connector 428 of the frame arm 416 (e.g., in preparation for precise electrical interface 828 of the connectors 432, 428).

[0156] In one arrangement, mechanical alignment 812 begins by first aligning the frame adapter 812 through a forward portion 744 of an alignment barrel 740 of a FRU 412 (e.g., of an adapter 700 that is secured to the FRU 412). 12 ), the alignment pin 429 and the alignment barrel 740 may be configured to be spaced 816 from the frame arm connector 428 and the FRU connector 432, respectively, by at least a substantially common distance (e.g., in the width dimension 405, see FIG. 12 ). With reference to FIGS. 13 , 16 , and 18 , for example, the tip portion 437 of the alignment pin 429 may be configured to enter the receiving bore 752 of the alignment barrel 740 via the forward portion 744 of the alignment barrel 740 upon the FRU 412 advancing 808 within the receiving bay 410. For example, the alignment pin 429 and the alignment barrel 740 may be configured and designed to be at a substantially common distance from a common reference line (or other reference location) of the outer rail member 450 in each of the width and height dimensions 405, 407 (see FIG. 12 ), such that the tip portion 437 can be substantially easily received within the alignment barrel 740.

[0157] Thereafter, continued advancement 820 of the FRU 412 toward the rear portion of the receiving bay 410 may serve to more completely and precisely align the connectors 428, 432 (e.g., corresponding pins / contacts of the connectors 428, 432) in both the width and height dimensions 405, 407. For example, camming action between the tapered tip portion 437 and the outer periphery of the receiving bore 752 adjacent the forward portion 744 of the alignment barrel 740 during continued advancement 820 may tend to correct for any differences in tolerances between the various components, such as the rail assembly 401, frame arm 416, FRU 412, adapter 700, etc., to precisely align the alignment pins 429 within the receiving bore 752, thereby precisely aligning the connections 428, 432 (e.g., due to the common distances discussed above). In one arrangement, the enlarged head 435 of the alignment pin 429 (labeled in FIG. 13 ) may have an outer diameter substantially the same as the inner diameter of the receiving bore 752 between the forward portion 744 and the rearward portion 748 of the alignment barrel 740 (e.g., such that there is a slight friction fit between the enlarged head 435 and the receiving bore 752 as the alignment barrel 740 moves relative to the alignment pin 429) to maintain the precise alignment discussed above that is primarily achieved through engagement between the tip portion 437 and the outer periphery of the forward portion 744 of the alignment barrel 740.

[0158] More specifically, configuring the outer diameter of the enlarged head 435 and the inner diameter of the receiving bore 752 to be substantially the same limits relative movement between the alignment pins 429 and the alignment barrels 740 (and thus between the connectors 428, 432) in the width and height dimensions 405, 407 as the FRU 412 advances (e.g., along the depth dimension 409) toward the rear portion of the receiving bay 410 to a fully loaded position. Although only a single set of alignment pins 429 and alignment barrels 740 are shown in the figures, it will be readily appreciated that more than one or more sets thereof may be included to further facilitate precise alignment of the connectors 428, 432. By way of example only, two or more alignment pins 429 may be received within respective apertures 431 in the wall of the housing 418 of the frame arm 416, while two or more respective alignment barrels 740 may be included in the mounting member 728 of the adapter 700.

[0159] In one arrangement, the connectors 428, 432 may have one or more alignment posts 475 and alignment openings 476 (labeled in FIGS. 13 and 16 ), with each alignment post 475 configured to be received in a respective corresponding alignment opening 476 after at least one alignment pin 429 is at least initially / partially received in a corresponding alignment barrel 740. In this regard, each alignment pin 429 / alignment barrel 741 combination may be considered a “primary” alignment feature (e.g., serving to provide initial alignment of the pins / contacts of the corresponding connectors 428, 432), while each alignment post 475 / alignment opening 476 combination may be considered a “secondary” alignment feature (e.g., serving to provide finer alignment of the pins / contacts of the corresponding connectors 428, 432). In any case, the method 800 may include electrically interfacing 828 the connectors 428, 432 after or during the mechanical alignment 812 of the connectors 428, 432. See FIG. In this regard, as all previously discussed, the FRUs 412 may be substantially automatically and / or seamlessly discovered by a frame center (e.g., frame center 120 of FIG. 1) and / or frame manager (e.g., frame manager 168 of FIG. 1) of a computing rack via a fixed interconnection topology 424, the frame center and / or frame manager may perform OOB management of the FRUs 412, FRU hot swap / removal requests may be generated by the frame arm 416 (e.g., upon pressing of a corresponding button on the frame arm 416) and passed to the frame center for processing, etc. One or more of the FRUs 412 of the rack 404 of FIG. 12 may be inserted into and removed from their respective receiving bays 410 of the rack 404 and each may join or leave the management network / context of the frame 400 in the manner discussed above using multiple substantially identical adapters 700, even if the multiple FRUs 412 collectively include different form factors (e.g., 1U and 2U FRUs), different functions, etc.

[0160] To help maintain firm and consistent electrical contact between the pins / contacts of the connectors 428, 432 when the FRU 412 is fully loaded into the receiving bay 410, the enlarged head 435 (e.g., locking portion) of the alignment pin 429 may interact and / or engage 824 with a locking portion in the form of a lip or flange 753 adjacent the rear portion 748 of the alignment barrel 740 to limit premature separation of the alignment pin 429 and alignment barrel 740 (and thus the connectors 428, 432). More specifically, the outer diameter of the enlarged head 435 may be at least slightly larger than the inner diameter of the flange 753, so that the flange 753 will need to be pushed at least slightly past the enlarged head 435 as the FRU 412 is advanced to the fully loaded position. In other words, the user may need to apply a slightly greater force to the front of the FRU 412 to move the flange 753 from a first side of the enlarged head 435 to the opposite side of the enlarged head 735. Once the flange 753 is pushed just past the enlarged head 435 (e.g., the flange 753 is moved just to the opposite side of the enlarged head 435), the connectors 428, 432 may be configured to fully engage / seat relative to one another. In this regard, the engagement between the flange 753 and the enlarged head 435 may act to resist relative movement between the alignment pin 429 and the alignment barrel 740, and thus between the connectors 428, 432 (in a direction that would otherwise tend to disengage the connectors 428, 432, e.g., due to vibration, shock, etc.), when a user does not intend to disengage the connectors 428, 432 (e.g., during a FRU hot swap operation).

[0161] Although engagement between the enlarged head 435 of the alignment pin 429 and the flange 753 of the alignment barrel 740 has been discussed as a manner of ensuring electrical engagement between the connectors 428, 432, other arrangements for doing so are contemplated and encompassed herein. For example, the flange 753 of the alignment barrel 740 may be configured to snap or otherwise fit into a groove or the like disposed about the circumference of the alignment pin 429. Additionally, the enlarged head 435 (or other feature) and the flange 753 (or other feature) may be disposed in locations on the alignment pin 429 and alignment barrel 740 other than those shown in the figures. For example, the enlarged head 435 may be disposed in a mid-portion of the alignment pin 429, and the flange 753 may be disposed inside a receiving bore 752 in the mid-portion of the alignment barrel 740.

[0162] 21-22, another embodiment of an adapter 900 is presented that is configured to allow insertion of a FRU 1012 into a receiving bay of a computing rack (e.g., receiving bay 410 of computing rack 404 of FIG. 12) with a blind mating interface between the frame arm 1016 of the receiving bay of the computing rack and the adapter 900, and is electronically discovered in a substantially seamless manner by management services software of the computing rack. 4 or 5. The adapter 900 may or may not be utilized in a rack in which the adapter 700 is utilized with a corresponding frame arm 416. Generally, the adapter 900 includes first and second mounting portions 9041, 9042 (e.g., each including one or more brackets) configured to be fixedly secured to opposing side portions 10131, 10132 of the FRU 1012, respectively, and a mounting portion 908 (e.g., one or more brackets) interconnecting the first and second mounting portions 9041, 9042. The mounting portion 908 includes at least a first connector 10321 (e.g., the blind mating connector 132 of FIG. 4 or the blind mating connector 132 of FIG. 1). 12. The first connector 10321 is connected to a receiving bay (not shown, but FIG. 12) of a computing rack when the FRU 1012 is inserted into the receiving bay. 4 or 428 in FIG. 13 ) adjacent to the rear of the receiving bay 410 of the rack 404 in FIG. The device is configured to allow electrical interfacing.

[0163] In one arrangement, the first and second mounting portions 9041, 9042 may each include at least a first mounting member 9121, 9122 (e.g., a plate, a flange, etc.) configured to be immovably secured (e.g., directly or indirectly) to the first and second side portions 10131, 10132 of the FRU 1012. For example, the first mounting members 9121, 9122 may each be rigidly secured (e.g., via fasteners, nuts, and aligned holes, not labeled) to an inner portion of a pair of inner rail members (not labeled) of the same rail assembly 1001 to which the respective frame arms 916 are secured (e.g., the frame arms 916 are secured to a pair of outer rail members 10501, 10502 of the rail assembly 1001, and the first mounting members 9121, 9122 are secured to a pair of inner rail members of the same rail assembly 1001). In this regard, the rigid mounting of the inner rail members to the side portions 10131, 10132 of the FRU 1012 (e.g., similar to how the inner rail members 454 may be rigidly secured to the side portions 413 of the FRU 412, as discussed above) serves to immovably or rigidly secure the adapter 900 to the FRU 1012. In another arrangement, the mounting portions 9041, 9042 may additionally or alternatively include additional mounting members (e.g., plates, flanges, etc., not shown) configured to be rigidly secured to the side portions 10131, 10132 of the FRU 1012 for increased stability of the system.

[0164] 21-22, the mounting portion 908 may include a mounting member 928 (e.g., a plate, bracket, flange, etc.) that rigidly interconnects the first and second mounting portions 9041, 9042 including at least a first receiving aperture (not shown, but similar to receiving aperture 732 in FIG. 14, albeit horizontally disposed in the embodiment of FIGS. 21-22) extending therethrough. The first receiving aperture is adapted to contact a forward portion of a first connector 10281 of a corresponding frame arm 1016 upon insertion of the FRU 1012 into a receiving bay of a rack (e.g., one of the receiving bays 410 of rack 404). (not shown) of the first connector 10321. 10321. It is noted that Figures 21-22 show the adapter 900 engaged to the frame arm 1016.

[0165] For example, the first connector 10321 snaps past the inner wall of the first receiving aperture. 908, thereby connecting the first connector 10321 to the mounting portion 908 and thus to the F The mounting portion 904 may include one or more flexible prongs, clips, or the like (not shown) adapted to lock or immovably secure to the RU 1012 (i.e., when the mounting portion 904 is fixed to or at least immovably secured relative to the FRU 1012). However, other means may be used to secure the first connector 10321 to the mounting portion 908. Embodiments are also contemplated and encompassed herein. In one arrangement, the mounting member 928 is The first and second mounting portions 9041, 9042 may be rigidly secured to the first mounting members 9121, 9122 in any suitable manner (e.g., threaded connections, welding, etc.). In another arrangement, a single piece of material may be appropriately formed and / or shaped to create the first and second mounting portions 9041, 9042 and the attachment portion 908.

[0166] The first and second mounting portions 9041, 9042 collectively space the mounting portion 908 a distance 924 from the rear portion 1014 of the FRU 1012 to advantageously provide a power port, a serial port, and a network port (not labeled) adjacent the rear portion 1014 of the FRU 1012 with a first connector 10321 secured to the mounting portion 908 of the adapter 900. Space may be provided for one or more cables, wires, etc. (not shown) to electrically connect to rear portion 1033. In one arrangement, adapter 900 may include a base or tray 960 that is suitably secured to first and second mounting portions 9041, 9042 and / or attachment portion 908. In either case, first and second mounting portions 9041, 9042 may be secured to attachment portion 908. 042 and mounting portion 908 (and tray 960, if included) are FRU 10 12 and the first connector 10321 adjacent to the rear portion 1014 of the Together they may form at least a partial housing for containing a number of interconnecting cables / wires / etc.

[0167] In one arrangement, the mounting member 928 engages a forward portion (not shown) of the second connector 10282 of the frame arm 1016 upon insertion of the FRU 1012 into a receiving bay of the rack. The second connector 10322 is attached to the housing 103 so that a front portion (not shown) of the second connector 10322 faces the 322 (e.g., blind mating connector) The adapter 900 may include at least a second receiving aperture (not shown) for receiving an electrical signal from the FRU 1012. Providing the second connector 10322, 10282 advantageously enables a blind-mating electrical connection between one or more additional ports of the FRU 1012 and additional cables / wires / networks, such as a rack. In one arrangement, one or more cables / wires (not shown) may be electrically connected between a high-speed network port 970 (e.g., Ethernet, RJ45) of the FRU 1012 and the second connector 10322 of the adapter 900. For example, the high-speed network port 970 may be utilized for actual data (e.g., signals) communication between the FRU 1012 and other devices and / or processes (e.g., other FRUs 1012 / 412 / 112 in the rack 404 / 104, devices or processes external to the rack 404, etc.), as opposed to for management communications between the frame center (e.g., frame center 120), frame arms 116 / 416 / 1016, FRUs 112 / 412 / 1012, etc.

[0168] The frame arm 1016 may generally have a form factor of width 405 and height dimensions 407 (labeled in FIG. 12) that substantially match the adapter 900. More particularly, the frame arm 1016 may include a housing 1018 (e.g., housing 118) that may be immovably secured in any suitable manner to a pair of outer rail members 10501, 10502 of a particular receiving bay's rail assembly 1001 in which the adapter 900 and FRU 1012 are translationally mounted. By way of example only, the housing 1018 may include first and second mounting portions 11041, 11042 that include a respective pair of grooves 1019 (see FIG. 23) configured to receive a corresponding pair of flanges (not shown) of the outer rail members 10501, 10502. In this regard, the housing 1018 can be slid along or relative to the outer rail members 10501, 10502 to a desired position (e.g., adjacent a rear portion of a receiving bay) and then non-movably secured to the outer rail members 10501, 10502 in any suitable manner (e.g., via tightening bolts (not shown) extending through opposing side portions of the housing 1018 relative to the outer rail members 10501, 10502). Other manners of securing the housing 1018 to a rack adjacent a rear portion of the receiving bay are also contemplated and encompassed within the scope of this disclosure.

[0169] The housing 1018 may also include a base or tray 1090 that is securely secured in any suitable manner between a first mounting portion 11041 and a second mounting portion 11042 to which the PCB 1022 (e.g., PCB 122, 422) may be secured, such as via passing fasteners 1017 through apertures (not shown) in the PCB 1022 into apertures (not shown) in the tray 1090 of the housing 1018. The PCB 1022, like the PCB 122 of FIG. 4 and the PCB 422 of FIG. 13, includes a memory that stores location data of the frame arm 1016 within the frame (e.g., frame 400) and a serial-I 2 C conversion logic, LED, etc. As shown, the first connector 10281 may have an optional 10321 of the adapter 900. The adapter 900 may be electrically connected to and secured to the PCB 1022 in any suitable manner to align with and electrically interface with the first connector 10321 of the adapter 900 to provide a fixed, inter-connector A connection topology (e.g., fixed interconnect topology 424) may facilitate electrical connections between each FRU 1012, the frame center, and one or more PDUs. More specifically, at least one I 2 C Cable or wire (for example, Figure 4 I 2 C line 20) is between the frame center and PCB1022 (PCB1022 I 2 C data bus), and at least one network cable A cable or wire (e.g., network wire 24 in FIG. 4) is connected between the frame center and a pin / contact (not shown) adjacent the rear portion 1070 of the first connector 10281. At least one power line or cable (e.g., power lines 28, 32 in FIG. 4) may be electrically connected (and thus may “pass through” PCB 1022) and may be electrically connected between one or more PDUs (e.g., PDUs 508, 512) and pins / contacts adjacent the rear portion 1070 of connector 1028.

[0170] A second connector 10282 is also connected to the frame arm 1016 (e.g., to the tray 1090). ) may be secured in any suitable manner to align with the second connector 10322 of the adapter 900. and may be adapted to electrically interface with the respective FRUs 1012 to facilitate electrical connection (e.g., actual data communication) between the respective FRUs 1012 and one or more networks via a fixed interconnect topology (e.g., fixed interconnect topology 424) and / or other cables / wires. For example, one or more network (e.g., Ethernet) cables (e.g., not shown) may be suitably electrically interconnected to a rear portion of the second connector 10282, thereby providing a Upon interfacing with the connectors 10322, 10282, an electrical (eg, data) connection is established between the FRU 1012 and one or more networks (eg, the Internet, a WAN, a LAN).

[0171] In one arrangement, the housing 1018 may additionally include a corresponding mounting portion 1120 configured to substantially abut the mounting portion 908 of the adapter 900 when the FRU 1012 is fully inserted into the receiving bay. The mounting portion 1120 includes respective first and second receiving apertures (not shown) for receiving and mounting (e.g., via clips or the like) of the first and second connectors 10281, 10282, respectively. In this regard, the adapter 900 and frame arm 1016 may be substantial mirror images of one another (e.g., the adapter 900 and frame arm 1016 have respective trays 960, 1090, mounting portions 908, 1120, first and second mounting portions 9041, 9042 and 11041, 11042, etc.).

[0172] In one embodiment, the frame arm 1016 may include at least one mechanical connector or alignment component, such as first and second alignment pins 10291, 10292 configured to be received in respective corresponding mechanical connectors, or alignment components, such as first and second alignment barrels (not shown, but similar to alignment barrel 740 of FIG. 16 ) of the adapter 900 of the FRU 1016. For example, the first and second alignment pins 10291 , 10292 are adjacent to the first and second connectors 10281, 10282 (for example, 13-19 , while each of the first and second alignment barrels may be disposed adjacent (e.g., directly adjacent, spaced adjacent, etc.) to the first and second connectors 10321, 10322, respectively. As discussed above in connection with the alignment pins 429 and alignment barrels 740 of FIGS. 13-19 , receipt of each of the alignment pins 10291, 10292 into the respective first and second alignment barrels upon insertion of the FRU 1012 into the receiving bay toward the frame arm 1016 facilitates precise alignment between the pins / contacts of the respective connectors 10281 / 10321, 10282 / 10322.

[0173] Where connectors 10281 / 10321, 10282 / 10322 have respective alignment posts and apertures (not labeled, but similar to alignment posts / apertures 475, 476 in FIGS. 13 and 16), each alignment pin 1029 / alignment barrel 940 combination may be considered a “primary” alignment feature (e.g., serving to provide initial alignment of the pins / contacts of the corresponding connectors 10281 / 10321, 10282 / 10322), while each alignment post / alignment aperture combination serves as a “primary” alignment feature (e.g., serving to provide initial alignment of the pins / contacts of the corresponding connectors 10281 / 10321, 10282 / 10322). A "secondary" aligner (which serves to provide finer alignment of the 322 pins / contacts) In some embodiments, the housing 1018 of the frame arm 1016 may include a cover 1061 (shown in FIG. 23; removed from FIGS. 21-22 for clarity) attached to the first and second mounting portions 11041 / 11042 and / or a mounting portion 1120 for protecting the PCB 1022 and the connectors 10281, 10282 and increasing the rigidity of the frame arm 1016, etc.

[0174] 20 may be applicable to the adapter 900 / frame arm 1016 combination in a manner similar to that discussed above in connection with the adapter 700 / frame arm 416 combination. For example, method 800 may involve connecting ends (not labeled) of a pair of inner rail members of FRU 1012 to a pair of outer rail members 10501 of a receiving bay (e.g., receiving bay 410) of a computing rack (e.g., computing rack 404). , 10502 into the channel (not labeled) 804 and the rear portion of the receiving bay The frame arm 1016 may include advancing 808 (e.g., sliding, moving, etc.) the FRU 1012 toward the frame arm 1016, mechanically aligning 812 the electrical connectors 10321, 10322 of the FRU 1012 with corresponding electrical connectors 10281, 10282 of the frame arm 1016 (e.g., via receiving alignment pins and / or posts in alignment barrels and / or openings), and electrically interfacing 828 the connectors 10281 / 10321, 10282 / 10322 after or during the mechanical alignment 812 of the connectors 10281 / 10321, 10282 / 10322. In this regard, as all previously discussed, the FRUs 1012 may be substantially automatically and / or seamlessly discovered by a frame center (e.g., frame center 120 of FIG. 1) and / or a frame manager of a computing rack (e.g., frame manager 168 of FIG. 1) via a fixed interconnection topology 424, the frame center and / or frame manager may perform OOB management of the FRUs 1012, and FRU hot swap / removal requests may be generated by the frame arm 1016, for example (e.g., when a corresponding button on the frame arm 1016 is pressed) and passed to the frame center for processing, etc.

[0175] In one arrangement, the adapter 900 and frame arm 1016 may be utilized for a “1U” FRU 1012 as the adapter 900, where the frame arm 1016 orients the various connectors 1032, 1028 along the width dimension 405 of the rack 404 (e.g., the longest dimension of the connectors 1032, 1028) (i.e., the adapter 700 and frame arm 404 of FIGS. 12-19). 16) (as opposed to along height dimension 407). For example, adapter 900 and frame arm 1016 may be utilized for a 1U FRU 1012, while adapter 700 and frame arm 416 may be utilized for a 2U FRU 412 that may or may not be utilized in the same rack 404. However, it is contemplated that each adapter 900 / frame arm 1016 and adapter 700 / frame arm 416 combination may be utilized with FRUs of other form factors (e.g., 1U, 2U, 3U, etc.).

[0176] In some circumstances, the second connector 10322 of the adapter 900 and the FRU 101 As a result of the electrical connection of the two high-speed network ports 970 (e.g., Ethernet, RJ45), electromagnetic interference (EMI) may be generated. In this regard, the enclosure created by the first mounting portion 9041, the second mounting portion 9042, and the mounting portion 908 may act as a "cage" that acts to attenuate or otherwise control the generated EMI. Additionally, one or more portions of the adapter 900 and / or the frame arm 1016 (e.g., tray 960, 1090; mounting portion 908, 1120; etc.) may include multiple perforations therethrough, etc., to facilitate air circulation, weight reduction, etc. It will be readily appreciated that numerous additions and / or departures may be made from the specific embodiments disclosed herein without departing from the spirit and scope of the invention. The illustrations and discussions herein are provided solely to aid the reader in understanding various aspects of the present disclosure. For example, although frame center 120 is disclosed as generally facilitating communication between FRUs 112 (e.g., as a switch) in connection with out-of-band management, another switch (e.g., InfiniBand) may be provided to facilitate communication between FRUs 112 and devices or networks external to frame 100. That is, FRUs 112 may be interconnected to frame center 120 (e.g., via interfaces of connectors 128, 132) for OOB management of FRUs 112, but FRUs 112 may be suitably interconnected to other types of switches for InfiniBand or other types of communication and / or data transfer (e.g., videoconferencing traffic). In alternative embodiments, more than one frame manager may be present within the frame. For example, frame center 120 may run a frame manager for OOB management of FRUs (e.g., in connection with power-on, hot-swapping, etc.), but one of the FRUs 112 may run another frame manager that administers higher level management tasks within frame 100.

[0177] In one arrangement, the FRU 112 / 412 may be properly locked in its particular receiving bay in the rack 104 / 404 and thus unable to be unlocked and removed until the service processor 106 and / or frame manager 168 have performed any necessary OOB management routines. In another arrangement, the service processor 160 and / or frame manager 168 may be configured to automatically initiate an OOB management routine (e.g., a take offline process) upon detecting a pull or tug on the FRU 112 / 412 in an attempt to remove the FRU 112 / 412 (e.g., as an alternative to pressing a button 182 on the corresponding frame arm 116 / 416). In this regard, the FRU 112 / 412 may be locked in its receiving bay of the rack 104 / 404 until the OOB management routines are completed.

[0178] Additionally, various combinations of one or more of the above-discussed arrangements and embodiments are contemplated. For example, FIG. 11 shows first and second frames 400, 404 each in the form of frame 100, but with at least one of the first and second frames 400, 404 in the form of frame 100. One of the FRUs may be in the form of a frame 100′ in FIG. 10 whereby a FRU 112′ in the form of a chassis enclosure contains a plurality of FRUs 112″ in the form of blade servers. As another example, the mounting portion 708 of the adapter 700 may have one or more additional receiving apertures for receiving additional connectors 432 which may be configured to electrically interface with one or more additional corresponding connectors 428 of the frame arms 416. As a further example, the adapter 900 and frame arms 1016 may be utilized for FRUs of other sizes (e.g., 2U, 3U, etc.).

[0179] The embodiments disclosed herein can be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer-readable medium for execution by or controlling the operation of a data processing apparatus. For example, the frame manager 168 may be provided in such a computer-readable medium, such as the frame center 120, one of the FRUs 112, and executed by a processor or processing engine (not shown) of the FRU 112. As another example, logic or software of the frame center 120 responsible for accessing the manager PROM image 148 and routing communications within the frame 100 may be provided in such a computer-readable medium (e.g., memory 144) of the frame center 120 and executed by a processor or processing engine (not shown) of the frame center 120 (not shown). The computer-readable medium can be a machine-readable storage device, a machine-readable storage substrate, a non-volatile memory device, a device affecting a machine-readable propagated signal, or a combination of one or more thereof. In this regard, the frame 100 may encompass one or more apparatuses, devices, and machines for processing data, including, by way of example, a programmable processor, a computer, or multiple processors or computers. In addition to hardware, the frame 100 may include code that creates an execution environment for the computer program in question, such as code that constitutes, for example, processor firmware, a protocol stack, a database management system, an operating system, or a combination of one or more of these.

[0180] Computer programs (also known as programs, software, software applications, scripts, or code) used to provide any of the functionality described herein (e.g., managing FRUs 112, routing communications, etc.) can be written in any suitable form of programming language, including compiled or interpreted languages, and can be deployed in any form, including as a stand-alone program, or as a module, component, subroutine, or other unit suitable for use in a computing environment. Computer programs do not necessarily correspond to files in a file system. A program can be stored in part of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple linked files (e.g., a file that stores one or more modules, subprograms, or portions of code). A computer program can be deployed to run on one computer or on multiple computers located at one site or distributed across multiple sites and interconnected by a communication network.

[0181] The processes and logic flows described in this specification may be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows may be implemented by special purpose logic circuitry, such as, for example, an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit). The apparatus may also be implemented as special purpose logic circuitry. Processors suitable for the execution of a computer program may include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor receives instructions and data from a read-only memory or a random access memory or both. In general, the elements of a computer are one or more processors for executing instructions and one or more memory devices for storing instructions and data. The techniques described herein may be implemented by a computer system configured to provide the described functionality.

[0182] Although this specification contains many details, these should not be construed as limitations on the scope of the disclosure or what may be claimed, but rather as descriptions of features specific to particular embodiments of the disclosure. Moreover, certain features described in this specification in the context of separate embodiments can also be realized in combination in a single embodiment. Conversely, various features described in the context of a single embodiment can also be realized in multiple embodiments separately or in any suitable subcombination. Furthermore, although features may be described above as acting in a certain combination, and even initially claimed as such, one or more features from a claimed combination can in some cases be deleted from the combination, and the claimed combination may be directed to a subcombination or a variation of the subcombination.

[0183] Similarly, although operations are depicted in the figures in a particular order, this should not be understood as requiring such operations to be performed in the particular order or sequentially shown, or that all of the depicted operations be performed to achieve a desired result. In some situations, multitasking and / or parallel processing may be advantageous. Furthermore, the separation of various system components in the above-described embodiments should not be understood as requiring such separation in all embodiments, and the described program components and systems may generally be integrated together in a single software and / or hardware product or packaged in multiple software and / or hardware products.

[0184] The above-described embodiments, including preferred embodiments and the best mode of the invention known to the inventors at the time of filing, are given as illustrative examples only.

Claims

1. 1. A system for automatically interconnecting a field replaceable unit (FRU) to an out-of-band (OOB) management network of a computing rack, the computing rack including a plurality of receiving bays for receiving a plurality of FRUs, the system comprising: a rail member configured to slide against a complementary rail member adjacent one of the plurality of receiving bays of the computing rack, the rail member including a plurality of mounting components for fixedly securing the rail member to a side portion of the FRU, the system further comprising: an adapter secured to a rail member for blindly connecting the FRU to the OOB management network of the computing rack, the adapter comprising: a mounting bracket fixedly secured to the rail member; a mounting member extending from the mounting bracket, the mounting member configured to receive a first blind mating connector configured to interface with an auxiliary second blind mating connector secured adjacent to a rear of the one of the plurality of receiving bays when the rail member slides relative to the auxiliary rail member.

2. The system of claim 1 , wherein the mounting member includes a first opening therein sized to receive the first blind mating connector.

3. The system of claim 2 , wherein the mounting member includes a second opening therein sized to receive an alignment pin secured adjacent the second blind mating connector.

4. The system of any of claims 1 to 3, wherein the mounting bracket includes a first mounting member and the adapter further includes a second mounting member for fixed attachment of the side portion of the FRU.

5. The system of claim 4 , wherein the second mounting member is spaced from the rail member.

6. 13. A method comprising: inserting a first FRU into a first receiving bay of a computing rack and electrically interconnecting the first FRU to an OOB management network of the computing rack using a first system as described in claim 1.

7. 7. The method of claim 6, further comprising inserting a second FRU into a second receiving bay of the computing rack and electrically interconnecting the second FRU to the OOB management network of the computing rack using the second system of claim 1, wherein the first and second systems are substantially identical and the first and second FRUs differ in at least one respect.

8. The method of claim 7 , wherein the at least one aspect includes at least one of a form factor, a power requirement, or a cooling requirement.

9. 1. A field replaceable unit (FRU) adapted for blind-mating electrical connection into a receiving bay of a computing rack, comprising: a housing having opposed top and bottom portions, opposed first and second side portions, and opposed front and rear portions; first and second slide rail members secured to the first and second side portions of the housing, respectively, each of the first and second slide rail members configured to interface with a complementary slide rail member secured adjacent a side portion of the receiving bay; and an adapter bracket having a mounting portion and a mounting portion, the mounting portion being fixedly secured to the housing and the mounting portion being spaced from the rear portion of the housing; and a blind mate connector secured to the mounting portion of the adapter bracket; a plurality of cables, a first cable of the plurality of cables being electrically interconnected between the blind mating connector and a power port in the enclosure and a second cable of the plurality of cables being electrically interconnected between the blind mating connector and a communications port of the enclosure, the blind mating connector of the FRU being configured to interface with an auxiliary blind mating connector secured adjacent a rear of the receiving bay, while the first and second sliding rail members slide relative to auxiliary sliding rail members of the receiving bay to electrically interconnect the FRU to a management network of the computing rack; The mounting portion is secured to one of the first and second slide rail members, a field replaceable unit (FRU).

10. A method for automatically interconnecting a field replaceable unit (FRU) to an out-of-band (OOB) management network of a computing rack, the computing rack including a plurality of receiving bays, the method comprising: receiving a plurality of FRUs in the plurality of receiving bays; a plurality of mounting components included in a rail member configured to slide against a complementary rail member adjacent one of the plurality of receiving bays of the computing rack, the rail member being fixedly secured to a side portion of the FRU; and an adapter secured to the rail member for blind mating and electrically interconnecting the FRU to the OOB management network of the computing rack, the adapter comprising: a mounting bracket fixedly secured to the rail member; a mounting member extending from the mounting bracket, the mounting member configured to receive a first blind mating connector configured to interface with an auxiliary second blind mating connector secured adjacent to a rear of the one of the plurality of receiving bays when the rail member slides relative to the auxiliary rail member.

11. The method of claim 10, wherein the mounting member includes a first opening therein sized to receive the first blind mating connector.

12. The method of claim 11, wherein the mounting member includes a second opening therein sized to receive an alignment pin secured adjacent to the second blind mating connector.

13. A method according to any of claims 10 to 12, wherein the mounting member includes a first mounting member and the adapter further includes a second mounting member for fixed attachment of the lateral portion of the FRU.

14. The method of claim 13, wherein the second mounting member is spaced from the rail member.

Images