UPLINK CONNECTIVITY IN RING NETWORKS

Abstract

A procedure that includes: generating, by a frame link module (FLM), in a frame belonging to a group of frames connected in a ring network, an uplink detection packet comprising information regarding a status and an identifier of a standby uplink port of the FLM (502), wherein the standby uplink port is configured in a Virtual Local Area Network (VLAN) that is not used by other ports of the ring network, so that the standby uplink port does not receive packets from the other ports; Determining, by the FLM and based on a system description of a Link Layer Discovery Protocol (LLDP) packet received by the standby uplink port from a customer network accessing the group of frames in the ring network, that the standby uplink port has a connection to the customer network (504); forwarding the uplink discovery packet by the FLM to the standby uplink port via a Peripheral Component Interconnect (PCI) interface of the FLM, wherein the standby uplink port is configured to override the VLAN while forwarding packets received via the PCI interface (506); Sending the uplink discovery packet by the FLM to the customer network via the standby uplink port, the uplink discovery packet being directed to a unicast destination address of an owner FLM in the ring network (508); Monitoring the receipt of the uplink discovery packet from the customer network by the owner FLM via a currently active uplink port in the ring network (510); and Determining, by the owner FLM and based on the successful receipt of the uplink discovery packet, that the standby uplink port and a set of switches in the customer network between the standby uplink port and the currently active uplink port are correctly configured (512).

Description

BACKGROUNDIn the area of scalable computing resources, network connectivity between frames of a scalable computing resource may represent a primary communication path for the exchange of data between these frames. The data may represent inputs to computing processes (e.g., data or applications), outputs of computing resources (e.g., computing results), communications to coordinate distributed processes, and other types of data. A frame is a primary entity of the scalable computing resource with embedded management and scalable connections that can be extended to improve the capabilities of the scalable computing resource. The scalable computing resource may include multiple frames connected in a ring network, wherein the frames of the scalable computing resource represent nodes of the ring network. Each frame may include one or more compute blades. In some ring networks of scalable computing resources, the frames may be interconnected via redundant network modules, also called frame link modules (FLMs), in each frame.The ring network is a network topology in which each node communicates with exactly two other neighbor nodes so that all communication runs over the ring network. To control traffic and prevent network loops, some ring network implementations have a ring owner. The ring owner blocks traffic on a link to prevent loops. At a particular time, there may be only a single ring owner. Upon failure of a ring owner, a new ring owner is selected.A frame may include two FLMs. Each of the FLMs may have an uplink port, also referred to as uplink, for connecting to external network resources, e.g., network switches in customer networks. When configured in the ring network, a group of frames may have only a single active uplink to a customer network and multiple standby uplinks that may be used as backup. When two uplinks become active simultaneously, an error condition such as a loop may be caused in the network and the network may break down. Coordinating between the FLMs in a group of frames to ensure that only a single uplink is deemed active allows the group of frames to function properly. In the event of a failure of an active uplink, a new single active uplink (from a series of standby uplinks) may be selected if the failure is detected. To function properly, the active uplink should be connected to an external network resource, e.g., a network switch in a customer network, and the active uplink and the network switch should be configured properly. Improper hardware connections to the active uplink or a misconfigured active uplink or network switches in the customer network may result in loss of access to the computing resources in the frames.US 2019,0 335 386 A1 discloses a method for determining whether an uplink from a network uplink is connected to an 802.1D non-compliant switch or directly to another FLM uplink.US 11 171 810 B1 discloses techniques for detecting misconfigureds of an uplink or an external network device connected to the uplink.Against this background, it was the object of the present invention to at least partially overcome the disadvantages of the prior art and in particular to help avoid network misconfiguredities.This object is at least partially achieved by a method according to independent method claims 1 and 10. Embodiments of the invention are the subject of the dependent claims.BRIEF DESCRIPTION OF THE DRAWINGSSome embodiments of the present disclosure will be described with reference to the following drawings. FIG. 1 is a block diagram of an example computing infrastructure including multiple scalable computing resources, a customer VLAN, and a management VLAN for determining uplink connectivity. FIG. 2 is a block diagram of an example computing infrastructure having multiple scalable computing resources with uplinks connected to network devices in an external network for determining uplink connectivity. FIG. 3 shows an example format of an uplink discovery packet for uplink connectivity determination. FIG. 4 is a block diagram of an example computing infrastructure having multiple ring networks of frame scalable computing resources with uplinks connected to network devices in a customer network for determining uplink connectivity. FIG. 5 is a flow diagram of an example method for determining uplink connectivity. FIG. 6 is a flow chart of another example of uplink connectivity determination. FIG. 7 shows an example of a computing device equipped with computer instructions in which various examples described herein for determining uplink connectivity may be implemented.In the drawings, identical reference numerals designate similar, but not necessarily identical, elements. The figures are not necessarily to scale, and the size of some parts may be exaggerated to clarify the example shown. Moreover, the drawings include examples and / or embodiments consistent with the description; however, the description is not limited to the examples and / or embodiments illustrated in the drawings.DETAILED DESCRIPTIONA scalable computing resource may be designed as a Removable Infrastructure (CI) that provides computing, storage, and network functions. A system administrator may manage, compile, and scale the resources of the scalable computing resource. The frame is the basic infrastructure of a scalable computing resource, in which the resources for data processing, storage, network, cooling, power supply and scalability are combined. Each frame may provide an uplink service to an external network so that the compute blades in the frames may be accessed from the outside. The external network refers to a network outside the frame or group of frames, which may be part of a virtual local area network (VLAN). The external network may be a user / customer network from which a network administrator / user may provide, maintain, and maintain the computing resources in the frames.A ring network topology is typical of a group of frames. Each frame in the ring network may be configured with redundant network modules, referred to as frame link modules (FLM), that include a processing resource, such as a central processing unit (CPU), and a switch that provides redundant network connection to the compute blades and connectivity between the frames. Within a particular frame, one FLM is active and the other is in standby mode. Each FLM may send commands to enable / disable connections to other network switches via remote sockets. Each FLM may also restart the other FLM in the same frame (referred to as neighbor FLM). Each FLM in the ring is able to provide the uplink to the customer network, but only one uplink is referred to as active uplink for the entire group of frames in the ring and the other uplinks are referred to as standby uplinks. According to the International Telecommunication Union (ITU) standard protocol G.8032, the Proprietary Node of the Ring Protection Layer (RPL), also referred to as a ring owner or owner FLM for a ring, is statically selected. In some examples, the FLM selected as a ring owner" may be responsible for determining the active uplink to provide the ring connectivity.As discussed in more detail below, the ring network of frames may include a management network and a client network supported by the scalable computing resource, and may use VLANs. In ring networks of frames, a loop prevention protocol such as G.8032 may be used. The operational, administrative and maintenance functions and mechanisms for such ring networks may be managed based on Y.1731. The customer network may comprise a large number of network switches. In order for the frames in a ring to be accessible to a user connecting to any network switch of the customer network, the switch must be correctly configured and connected to the same layer 2 network (L2) to which the uplink ports of the ring are connected. L2 or data link layer refers to the second level of the Open Systems Interconnection (OSI) seven layer model of computer networks. This layer is the protocol layer that transfers data between the nodes of a network segment over the physical layer. For example, if a switch in the customer network is connected to L2 of a network other than the uplink ports, the computational resources in the ring may not be available from that particular switch. For example, if an uplink is connected to L2 of a different network than the other uplinks and customer switches, the computational resources in the ring may be unavailable from such an uplink and its connected neighbor switches in the customer network. Therefore, proper configuration and connection of the uplinks and switches in the customer network is required to be part of a single layer of a network, i.e. L2 of the network. The techniques described above pose problems with uplink configuration and connectivity to the customer network that are not accounted for in conventional techniques.Accordingly, the techniques disclosed below enable the determination of whether a standby uplink port and network switches in the customer network between the standby uplink port and the current active uplink port are correctly configured and connected to the same network layer, i.e., L2 of the network. This may reduce / eliminate loss of connectivity between the customer network and the computing resources in the ring, thus reducing downtime. Therefore, the presented techniques are more efficient compared to the conventional techniques.Methods for determining whether the uplinks and network switches in the customer network are correctly configured are disclosed. The method includes generating, by an FLM, an uplink discovery packet in a frame belonging to a group of frames connected in a ring network. The uplink discovery packet includes information regarding a status and an identifier of a standby uplink port of the FLM. The standby uplink port is configured in a virtual local area network (VLAN) that is unused by other ports of the FLM, so that no packets are forwarded from the other ports to the standby uplink port. The method includes determining, by the FLM and based on a system description of an LLDP packet received by the standby uplink port from a customer network that the standby uplink port has a connection to the customer network. The method includes forwarding, by the FLM, the uplink discovery packet to the standby uplink port via a peripheral component interconnect (PCI) interface coupled to the standby uplink port. The standby uplink port is configured to disable the VLAN while forwarding packets received over the PCI interface. The method comprises sending, by the FLM, the uplink discovery packet to the customer network via the standby uplink port, wherein the uplink discovery packet is directed to a unicast destination address of an owner FLM in the same ring network. The method includes monitoring reception of the uplink discovery packet from the customer network by the owner FLM over a current active uplink port in the ring network. The method includes determining, by the owner FLM and based on the successful reception of the uplink discovery packet, that the standby uplink port and a set of switches in the customer network between the standby uplink port and the current active uplink port are correctly configured. In one example, in response to determining that the standby uplink port and the group of switches are correctly configured, the standby uplink port may be included in a group of potential uplink ports from which a new active uplink port may be selected in the event of a failure of the current active uplink port. In this way, a misconnected uplink or an uplink with incompatible VLAN configuration is prevented from being selected as the active uplink, thereby reducing / reducing the risk of loss of connectivity to the frames in the ring network. This may reduce down times due to loss of access to the compute resources in the ring and improve performance. In another example, a misconfigured one of the switches in the customer network may be detected based on not receiving the uplink discovery packet. This may also allow identification of configuration or connection failures in the customer network.FIG. 1 is a block diagram of an example computer infrastructure 100 for determining uplink connectivity. In this example, the customer network 105 is connected to a group of frames (represented by Frame 1 (110) and Frame 2 (115)). In some examples, there may be more than two frames in a ring, but for simplicity only two are shown in this example. As indicated by arrow 120- 1, frame 1 may include a series of blades (B 1, B 2,... BN) and a Customizable Infrastructure (CI) module. Arrow 120- 2 indicates that frame 2 may be configured in a similar manner. Frame 1 also includes two frame link modules (FLMs), namely FLM 1 (140) and FLM 2 (145) (also referred to as a network module). Frame 2 also includes two FLMs, namely FLM 3 (150) and FLM 4 (155). These FLMs provide connectivity for the computing resources represented by the respective blades within their frame. Each of the blades is shown having a network connection to a network switch (referred to as network switch 160 in the case of FLM 1) located in each individual FLM (e.g., FLM 1(140) through FLM 4(155)). Each FLM also includes a processor (referred to as processor 165 for FLM 1) that facilitates configuration, monitoring, and maintenance of a corresponding network switch (referred to as network switch 160 for FLM 1). Thus, each frame with an FLM may represent a network node in a ring network and may include a processor, an uplink port for transmitting and receiving packets on an external network, and memory (not shown in FIG. 1 ) coupled to the processor and having instructions executable by the processor stored therein.The connection from the frame group (frame 1 and frame 2) to a customer network 105 is made via a single active uplink 125 from one of the multiple network switches connected in a ring network via the multiple FLMs of the frame group. That is, all communication outside the group of connected frames will be over uplink 125. It is noted that an "uplink" as used herein may also be referred to as a "management port" or "uplink port.". Users connecting to the customer network 105 may access the computing resources of the frames in the ring network. The uplink 125 is configured in a management mode in which the uplink 125 is set to exchange traffic with the customer network 105. The management mode is an indication of an uplink operation mode in which the uplink is connected to an external network such as the customer network 105 and the uplink can exchange data traffic with the external network. Typically, uplink 125 is connected to a port of a network switch (not shown in FIG. 1 ) in customer network 105. Both the uplink 125 and the port of the network switch in the customer network should be configured to receive untagled traffic and reject tagged VLAN traffic for operation in management mode. The standby uplinks 125- 2, 125- 3, and 125- 4 are shown as being available from other network switches (e.g., when needed due to a failure of the uplink 125). In one example, the standby uplinks are also configured in the administrative mode.As further illustrated in computer infrastructure 100, customer network VLAN 130 connects each of network switches 160 in an Ethernet ring network and extends to customer network 105. A second ring network, management VLAN 135, is also shown as an isolated network in computer infrastructure 100. The management VLAN 135 is shown in a thicker line than the customer network VLAN 130 and also connects each of the network switches. It is noted that if a group of frames are properly configured, each node is directly connected to each neighbor node (either in the same frame or in a neighboring frame) and there are no intervening network devices.A virtual LAN (VLAN) refers to a broadcast domain that is partitioned and isolated (i.e., logically isolated) in a computer network on the data transmission (OSI) layer 2. LAN is the abbreviation for Local Area Network (Local Area Network). In the context of a VLAN, "virtual" refers to a physical object that is re-created and altered by additional logic. A VLAN is a user-defined network that is created from one or more existing LANs. It enables the combination of device groups from a plurality of networks (both wired and wireless) to form a single logical network. The result is a virtual LAN that can be managed like a physical local area network, for example, the management VLAN 135 in the computer infrastructure 100.In FIG. 2, another view of FIG. 1 is shown as computer infrastructure 200, with certain elements not shown and certain additional elements shown for clarity of discussion. As in FIG. 1, computer infrastructure 200 shows four FLMs, namely FLM 1 (240), FLM 2 (245), FLM 3 (250) and FLM 4 (255). The four FLMs are connected in a management VLAN 235, similar to the management VLAN 135 of FIG. 1, thus forming a ring network of a group of FLMs. Each FLM is also shown to include a network switch 260- 1 to 260- 4 that is respectively coupled to a respective processor 265- 1 to 265- 4 that is coupled to a respective memory. In FIG. 2, memory 270- 2 is coupled to processor 265- 2 and memory 270- 4 is coupled to processor 265- 4. The memories associated with the processors 265-1 and 265-3 in FLM 1 and FLM 3, respectively, are not shown in Figure 2.Each of the processors 265 may include a single or multiple central processing units (CPU), or other suitable hardware processor or processors, such as a network ASIC. Each of the memories 270 may store machine readable instructions executable by the respective processors 265. Each of the memories 270 may include any suitable combination of volatile and / or nonvolatile memory, such as combinations of random access memory (RAM), read-only memory (ROM), flash memory, and / or other suitable memories.As shown in FIG. 2, the memory 270- 4 stores instructions to be executed by the processor 265- 4, including instructions for the discovery packet generator 280, the connection monitor 282, the interface manager 284, and the discovery packet transmitter 286. Further, the memory 270- 2 stores instructions to be executed by the processor 265- 2, including instructions for the discovery packet receiver 290 and the configuration checker 292.As further illustrated in FIG. 2, each network switch 260- 1 to 260- 4, also collectively referred to as network switches 260, has at least two ports connected to other network modules ( 240, 245, 250, and 255). As in FIG. 1, it may be desirable that each of the two link ports be directly connected to a network module within the ring topology and that there be no intermediate devices between the network modules (240, 245, 250 and 255). As in FIG. 1, frame 1 may include a series of blades (B 1, B 2,... BN) and a Customizable Infrastructure (CI) module. Frame 2 may be configured in a similar manner.In an example, each of the processors 265 may be connected to a corresponding network switch 260 via a peripheral component interconnect (PCI) interface (not shown). The PCI interface refers to a local bus for connecting hardware devices in the FLM and is part of the PCI local bus standard. The PCI bus supports the functions of the processors. Packets from the processors 265 over the PCI interface to the network switches 260 may be routed to any port connected to the network switches 260.Each of the network switches 260 has an uplink port. As shown in FIG. 2, uplink 225- 1 of network switch 260- 1 is connected to a switch 275- 1 in customer network 205. The uplink 225- 4 of the network switch 260- 4 is connected to the switch 275- 2 in the customer network 205. Similarly, uplinks 225-2 and 225-3 are connected to switches (not shown) in customer network 205. The customer network 205 is an external network to which the frame groups 1 and 2 can be connected via a single active uplink 225- 1, so that access to the compute blade resources and maintenance can be performed externally from the customer network 205. Customer network 205 may include a series of switches interconnected via Ethernet connections such that the switches including switches 275- 1 and 275- 2 are located in the same network layer, i.e., L2 of the network. The customer network 205 is similar to the customer network 105 of FIG. 1. In the illustration of FIG. 2, the uplink 225- 1 is illustrated as the only active uplink selected by an owner FLM, i.e., FLM 2, from the group of frames in the ring network. In some examples, the owner FLM, also referred to as a ring protection layer (RPL) owner node, is manually selected by a technician according to the ITU G.8032 standard. A configuration program (e.g., a user interactive program) may be provided to allow the technician to select a node in the ring and determine it as an owner FLM. Thus, the owner FLM may generally be selected by user intervention, either at the first startup or in response to an owner FLM failing.The owner FLM may be responsible for determining which FLM may enable its uplink to the customer network to enable ring connectivity. In some examples, the owner FLM may use a management port selection protocol to select the active uplink or to notify that a new uplink is needed. Other nodes may also use this protocol to notify the owner FLM that they have an uplink available (e.g., a registration process to "queue"). In one example, all FLMs in the ring that have connected an uplink port may send a packet to the owner FLM by notifying that they have an uplink and using it as an active uplink (e.g., registration with the owner FLM). The contents of this packet may contain the MAC address of the FLM with the uplink and the MAC address of the other FLM in the frame (i.e. its neighbor). The owner FLM may store these packets in a registration table and randomly select an FLM from the registration table to enable and become active its uplink while the other uplinks in the table are determined to be standby uplinks.In the event of a failure of an active uplink connection, a new active uplink connection (from a number of available backups) can be selected as soon as the failure is detected. Heartbeat messages may be used to exchange coordination and status information between the cooperating nodes. The FLM with the active uplink may send heartbeat status messages to the owner FLM, suggesting that its connection is active. If the owner FLM does not receive heartbeat messages from the active uplink FLM, the owner FLM may conclude that the active uplink FLM is experiencing an error or restarting. In response to determining that the FLM has failed with the active uplink, the owner FLM may randomly select one of the standby uplinks to become active from the registration table.In initial setup and configuration, all uplinks in the computer infrastructure 200 are configured by default in different VLANs, either manually or using application programming interfaces (APIs) and software utility programs. Each uplink port is marked for receiving data traffic in its assigned VLAN, so that the data traffic of the uplinks is separated by these VLANs. Once the owner FLM has determined an uplink to be the only active uplink, a technician may configure the single active uplink to be the untagled port. Thus, each standby uplink remains configured in a different VLAN in which no other ports of the ring network are configured. Thus, the standby uplinks are isolated in their respective VLANs, and no traffic is received at the standby uplinks from the other ports of the FLM or from the customer network, except link layer discovery protocol (LLDP) packets. In FIG. 2, uplinks 225- 2, 225- 3, and 225- 4 are labeled as standby uplinks.In FIG. 2, a technician may select FLM 2 as an owner FLM, as already explained. The uplinks 225- 1 to 225- 4 are configured in a management mode by default. The owner FLM 2 may select the uplink 225- 1 as currently active uplink port and the uplinks 225- 2 to 225- 4 as standby uplink ports. According to the techniques of the present disclosure, the owner FLM may determine whether a standby uplink port and a set of switches in the customer network between the standby uplink port and the current active uplink port are correctly configured. Although the manner in which this determination is made is explained with reference to FLM 4 and its standby uplink 225-4, the described method (optionally with corresponding changes) can also be applied to other FLMs, namely FLMs 2 and 3, with standby uplinks 225-2 to 225-3.In response to the selection of the single active uplink 225- 1 by the owner FLM 2 and the determination of 225- 2 to 225- 4 as standby uplinks, the discovery packet generator 280 may generate an uplink discovery packet including information regarding a status and an identifier of the standby uplink port 225- 4 of the FLM 4. The status of the standby uplink 225- 4 may indicate that the uplink 225- 4 is in the standby state, and the identifier of the standby uplink 225- 4 may include a MAC address of the standby uplink 225- 4. In one example, the uplink discovery packet includes a MAC address of the owner FLM, i.e., FLM 2, and interface information of a switch 275- 2 from the group of switches to which the standby uplink port 225- 4 is connected. In one example, the uplink discovery packet is directed to a unicast destination address of an owner FLM in the ring network, such as FLM 2 in FIG. 2. Further, the standby uplink 225- 4 may be configured in a VLAN (not shown) that is unused by other ports of the ring network, such that the standby uplink 225- 4 does not receive packets from the other ports.In one example, the uplink discovery packet may have a Ring Advanced Protection Switching (R-APS) Protocol Data Unit (PDU) format and transmitted with G.8032 protocol messages. An example of an R-APS PDU format is shown in Figure 3. In the packet shown in FIG. 3, version, opcode, and flags are defined by the Y.1731 standard for Ethernet-based networks operational, administrative, and maintenance functions and mechanisms (OAM). R-APS specific information is added in 32 octets from 5 to 36. Type-length-value (TLV) is a coding scheme used for optional information elements in a particular protocol. By means of TLV, information of the uplink discovery packet can be included in the R-APS PDU. Optional TLVs may be included in octet number 37. In the uplink discovery packet, the optional TLV may include a MAC address of the FLM from which the uplink discovery packet originates, i.e., in this case, FLM 4, the IPv6 link layer address of FLM 4, the bay number of FLM 4, the MAC address of the ring owner, i.e., the MAC address of FLM 2, the LLDP port description, the state of FLM 4, and timeout information. Referring to FIG. 2, the ring owner MAC address is the MAC address of FLM 2 selected as the owner FLM, the LLDP port description of the uplink discovery packet of FLM 4 may include the interface information of the switch 275- 2 to which the uplink 225- 4 of FLM 4 is connected. The state of FLM 4 may indicate that uplink 225- 4 of FLM 4 is in standby mode. The timeout information defines a time limit for which the information in the uplink discovery packet can be considered valid. After expiration of the time limit mentioned in the timeout information, the uplink determination packet can be considered as expired.The processor 265- 4 executes the link monitor 282, which may cause the FLM 4 to determine that the standby uplink port 225- 4 has a connection to the customer network 205. The presence of the connection may indicate that the standby uplink 225- 4 is connected to the switch 275- 2 in the customer network 205 via an Ethernet connection. In an example, the link monitor 208 may check whether an LLDP packet is received at the uplink port 225- 4. The link monitor 208 may use a packet sniffer or packet analyzer to recognize and analyze the LLDP packet. In one example, a system description may be present in the LLDP packet that may indicate a source device from which the LLDP packet originated. Based on the analysis of the system description associated with the LLDP packet, the source device of the LLDP packet may be identified. Thus, the FLM 4 may determine that the standby uplink port 225- 4 is connected to the source device, i.e., to the switch 275- 2 in the customer network 205 in this case. Therefore, the FLM can determine whether the standby uplink port has a connection to the customer network based on the system description of the LLDP packet received at the standby uplink port.The processor 265- 4 executes the interface manager 284, which may cause the FLM 4 to forward the uplink discovery packet to the standby uplink port 225- 4 via a Peripheral Component Interconnect (PCI) interface connected to the network switch 260- 4. Because the standby uplink is configured as a tagged port in its designated VLAN by default, the interface manager 284 may configure the standby uplink 225- 4 to override the VLAN in which it is configured while forwarding packets received over the PCI interface. Disabling the VLAN involves forwarding the uplink discovery packet to the customer network as untagled traffic even though the standby uplink 225-4 may be isolated in a VLAN. The processor 265- 4 executes the discovery message transmitter 286, which may cause the FLM 4 to transmit the uplink discovery packet to the customer network 205 via the standby uplink port 225- 4. In an example, the uplink discovery packet may be received at a port of the switch 275- 2 in the customer network 205 to which the uplink 225- 4 is connected. In one example, switch 275- 2 is a top-of-rack (ToR) switch and may be connected to other ToR switches in customer network 205, such as switch 275- 1. Top-of-rack switching is a data center architecture in which computing equipment such as servers, devices, and other switches located in the same or adjacent rack are connected to an in-rack network switch, also called a ToR switch. The uplink discovery packet transmitted over the standby uplink 225- 4 may be received at a port of the switch 275- 2 and travel through the switches in the customer network 205. Since the uplink discovery packet is directed to the owner FLM 2, if the switches in the customer network and the uplinks are correctly configured, the uplink discovery packet should be forwarded back to the ring network via the current active uplink 225-1 of FLM 1 to the destination, i.e., the owner FLM 2.The processor 265- 2 executes the discovery packet receiver 290 in the owner FLM 2, which may cause the owner FLM 2 to monitor receipt of the uplink discovery packet from the customer network 205 via the current active uplink port 225- 1 in the ring network. In one example, processor 265- 2 may execute discovery packet receiver 290 to check for the presence of an opcode in the 37th octet of the R-APS PDU format, as shown in FIG. 3. In one example, if the uplink discovery packet is successfully received at the owner FLM 2, the processor 265- 2 performs the configuration check 292, which may cause the owner FLM 2 to determine that the standby uplink port 225- 4 and the set of switches in the customer network 205 between the standby uplink port 225- 4 and the current active uplink port 225- 1 are correctly configured based on the successful reception of the uplink discovery packet. Although the above explanation relates to the standby uplink port 225- 4 of FLM 4, the configuration of the standby uplinks 225- 2 and 225- 3 may be similarly checked.In response to determining that the standby uplink 225- 4 and the set of switches are correctly configured in the customer network 205, the owner FLM 2 may store the status and identifier of the standby uplink 225- 4 in a set of potential uplink ports. The set of potential uplink ports may include a list of uplinks determined to be correctly configured. In one example, the set of potential uplink ports may be included in the registration table from which the owner FLM 2 may select a new active uplink in response to the failure of the current active uplink 225- 1.In some examples, the processor 265- 2 performs the configuration check 292, which may cause the owner FLM 2 to determine that the standby uplink port 225- 4 or at least one of the switches in the customer network 205 between the standby uplink port 225- 4 and the current active uplink port 225- 1 are misconfigured based on not receiving the uplink discovery packet. For example, if the uplink discovery packet is not received from the owner FLM 2 after a monitoring threshold time of, for example, 60 seconds, it may be determined that the uplink discovery packet has not been received. The threshold time may be manually set or predefined by a user. If it is determined that the standby uplink port 225- 4 or at least one of the switches in the customer network 205 is misconfigured, the FLM 2 may determine that the group of frames, i.e., frame 1 and frame 2, is unavailable from the at least one switch. Further, in one example, the owner FLM 2 may remove the information regarding the status and the identifier of the standby uplink port from the set of potential uplink ports when a threshold time has elapsed from the reception of the uplink discovery packet. In one example, the threshold time is set in the timeout information of the uplink discovery packet. The timeout information defines a time limit for which the information in the uplink discovery packet can be considered valid. After the time interval mentioned in the timeout information has elapsed, the uplink determination packet can be considered as expired. In one example, the timeout is specified by a network administrator and is a predefined value.FIG. 4 is a block diagram of an example computing infrastructure 400 having multiple ring networks of frame scalable computing resources with uplinks connected to network devices in a customer network. As in FIG. 2, computer infrastructure 400 shows four FLMs, namely FLM 1 (440), FLM 2 (445), FLM 3 (450) and FLM 4 (455). Also shown is each FLM including a network switch 460- 1 to 460- 4 that is respectively connected to a corresponding processor 465- 1 to 465- 4 that is connected to a corresponding memory. In FIG. 4, memory 470- 2 is coupled to processor 465- 2 and memory 470- 4 is coupled to processor 465- 4. The memories coupled to processors 465-1 and 465-3 in FLM 1 and FLM 3, respectively, are not shown in Figure 4. Each of the processors 465 and each of the memories 470 may be identical to the processors 265 and the memories 270 of FIG. 2.As shown in FIG. 4, the memory 470- 4 stores instructions to be executed by the processor 465- 4, including instructions for the discovery packet generator 480, the connection monitor 482, the interface manager 484, and the discovery packet transmitter 486. Discovery packet generator 480, connection monitor 482, interface manager 484, and discovery packet transmitter 486 are example implementations of discovery packet generator 280, connection monitor 282, interface manager 284, and discovery packet transmitter 286 of FIG. 2. moreover, memory 470- 2 stores instructions to be executed by processor 465- 2, including instructions for discovery packet receiver 490 and configuration checker 492. Discovery packet receiver 490 and configuration checker 492 are example implementations of discovery packet receiver 290 and configuration checker 292 of FIG. 2.As further illustrated in FIG. 4, each network switch 460- 1 to 460- 4, also collectively referred to as network switches 460, has at least two ports connected to other network modules ( 440, 445, 450, and 455). In one example, each of the processors 465 may be connected to a corresponding network switch 460 via a peripheral component interconnect (PCI) interface. The PCI interface refers to a local bus for connecting hardware devices in the FLM and is part of the PCI local bus standard. The PCI bus supports the functions of the processors. Packets from the processors 465 to the network switches 460 via the PCI interface may be routed to any port connected to the network switches.As shown in FIG. 4, uplink 425- 1 of network switch 460- 1 is connected to a switch 475- 1 in customer network 405. The uplink 425-4 of the network switch 460-4 is connected to the switch 475-2 in the customer network 405. Similarly, uplinks 425-2 and 425-3 are connected to switches (not shown) in customer network 405. Customer network 405 is similar to customer network 205 of FIG. 2, and may include a series of switches interconnected via Ethernet connections such that the switches including switches 475- 1 and 475- 2 are in the same layer, i.e., L 2 of the network.As shown in FIG. 4, the computer infrastructure 400 includes two ring networks, as opposed to a single ring network 235 in FIG. 2, and in FIG. 4, a first management VLAN 435 connects each of the network switches 460- 1 and 460- 2 in an Ethernet ring network. A second management ring network, second management VLAN 436, is also shown as an isolated network in computer infrastructure 400. The second management VLAN 436 is shown as a dashed line and connects the network switches 460- 3 and 460- 4. Although the first management VLAN 435 is shown in FIG. 4 as comprising a single frame.In one example, frames 1 may be connected multiple frames in a single management ring. Although the second management VLAN 436 is shown to include a single frame, i.e., frame 2, in one example, multiple frames may be connected in a single management ring. In the illustration of FIG. 4, it is to be noted that FLM 2 is selected as an owner FLM for the first management VLAN 435 and FLM 3 is selected as an owner FLM for the second management VLAN 436. Because two separate ring networks, the first management VLAN 435 and the second management VLAN 436, are illustrated, two separate ring owners are selected for each of the ring networks. Uplink 425-1 is shown as the active uplink selected by owner FLM 2 for the ring network of first management VLAN 435 and uplink 425-3 is shown as the active uplink selected by owner FLM 3 for the ring network of second management VLAN 436.According to the techniques of the present disclosure, it may be determined whether an uplink port and a set of switches in the customer network between the uplink port and the currently active uplink port are correctly configured. Although the following description refers to FLM 4 and its uplink 425-4, the described method can be applied to other FLMs, namely FLMs 1, 2 and 3, with the uplinks 425-1 to 425-3 with the required changes.The processor 465- 4 executes the discovery packet generator 480, which may cause the FLM 4 to generate an uplink discovery packet that includes information regarding a status and an identifier of the FLM's uplink port 425- 4. In one example, the uplink discovery packet includes a broadcast destination address. In one example, the broadcast destination address is a MAC address that indicates to a switch receiving the uplink discovery packet that it should forward the uplink discovery packet to all its ports except the port from which it was received. An example of a broadcast destination MAC address may be FF:FF:FF:FF:FF:FF. Thus, the uplink discovery packet, when received by the switch 475-2, may be flooded in the L2 of the network, including the switches 475-1 and 475-2 of the customer network 405 and the network switches 460 of the rings connected in the same layer of the network. Further, the uplink discovery packet may be transmitted from the network switches 460 to owner FLMs belonging to each of the rings, such as owner FLM 2 and owner FLM 3 of the first management VLAN ring and the second management VLAN ring, respectively. In an example, the uplink discovery packet includes an interface information of the switch 475- 2 of the set of switches to which the uplink port 425- 4 is connected.The status of the uplink 425- 4 may be "active" or "standby", indicating that the uplink 225- 4 is in the active or standby state, and the identifier of the uplink 425- 4 may include a MAC address of the uplink 425- 4. In one example, the uplink discovery packet may have the Ring Advanced Protection Switching (R-APS) Protocol Data Unit (PDU) format and may be transmitted using G.8032 messages. In some examples, the uplink discovery packet may have a format as shown in FIG. 3Processor 465- 4 executes link monitor 482, which may cause FLM 4 to determine that uplink port 425- 4 has a connection to customer network 405. The presence of the connection may indicate that uplink 425-4 is connected to switch 475-2 in customer network 405 via an Ethernet connection. In one example, processor 465- 4 may execute link monitor 482 to check whether an LLDP packet is received at uplink port 425- 4. When an LLDP packet is received at uplink port 425-4, processor 465-4 may execute link monitor 482 to analyze the LLDP packet. In one example, a system description may be present in the LLDP packet that may indicate a source device from which the LLDP packet originated. Based on the analysis of the system description associated with the LLDP packet, the source device of the LLDP packet may be identified. Thus, the FLM 4 may determine that the uplink port 425- 4 is connected to the source device, i.e., the switch 475- 2 in the customer network 405.The processor 465- 4 executes the interface manager 484, which may cause the FLM 4 to forward the uplink discovery packet to the uplink port 425- 4 via a PCI interface connected to the network switch 460- 4. The processor 465- 4 executes the discovery message transmitter 486, which may cause the FLM 4 to transmit the uplink discovery packet to the customer network 405 via the uplink port 425- 4. In one example, the uplink discovery packet is received at a port of the switch 475-2 in the customer network 405 to which the uplink 425-4 is connected. In one example, switch 475- 2 is a top-many rack (ToR) switch and may be connected to other ToR switches in customer network 405, such as switch 475- 1. The uplink discovery packet transmitted over uplink 425-4 may be received at a port of switch 475-2 and travel over the network of switches in customer network 405. Since the uplink discovery packet is directed to the broadcast destination address, if the network is correctly configured, the packet should be returned to the ring network via the respective active uplinks of the ring network. Referring to FIG. 4, the uplink discovery packet should be forwarded to the owner FLM 2 back to the first management VLAN ring via the active uplink 425-1. Also, since the uplink discovery packet is a broadcast packet, a copy of the uplink discovery packet should be forwarded to the owner FLM 3 back to the second management VLAN ring via the active uplink 425-3. The FLM owners in the rings check whether or not the packet has been received in the ring.The processor 465- 2 executes the discovery packet receiver 490 in the owner FLM 2, which may cause the owner FLM 2 to monitor receipt of the uplink discovery packet from the customer network 405 via the current active uplink port 425- 1 in the first management VLAN of the ring network. In one example, processor 465-2 executes discovery packet receiver 490 to check for the presence of an opcode in the 37th octet of the R-APS PDU format, as shown in FIG. 3. If the OpCode is present, the processor 465- 2 may execute the discovery packet receiver 490 to determine that the uplink discovery packet has been received. In one example, if the uplink discovery packet was successfully received at the owner FLM 2, the processor 465- 2 executes the configuration verifier 492, which may cause the owner FLM 2 to determine that the uplink port 425- 4 and the set of switches in the customer network 405 between the uplink port 425- 4 and the current active uplink port 425- 1 are correctly configured based on the successful reception of the uplink discovery packet. Although the above embodiments relate to the uplink port 425-4 of FLM 4, the configuration of the uplinks 425-1 to 425-3 may be similarly checked.In some other examples, processor 465- 2 executes configuration verifier 492, which may cause owner FLM 2 to determine that uplink port 425- 4 or at least one of the switches in customer network 405 between uplink port 425- 4 and the current active uplink port 425- 1 is misconfigured based on not receiving the uplink discovery packet. For example, when the uplink discovery packet is not received from the owner FLM 2 after a monitoring time of, for example, 60 seconds, it may be determined that the uplink discovery packet has not been received. If it is determined that the standby uplink port 425- 4 or at least one of the switches in the customer network 405 is misconfigured, FLM 2 may determine that the frame group, i.e., frame 1 and frame 2, is unavailable from the at least one switch.FIG. 5 illustrates an example method 500 for determining whether the uplinks and network switches in the customer network are correctly configured. The method 500 may be performed by FLMs in a ring network, such as the ring network represented by the first management VLAN 235 in FIG. 2. In some examples, steps 502- 508 may be performed by an FLM, e.g., FLM 2- FLM 4, each having a standby uplink, and steps 510- 512 may be performed by an owner FLM, e.g., owner FLM 2 of FIG. 2. method 500 may be performed as one or more instructions on a machine (e.g., by at least one processor), the one or more instructions being included on at least one machine-readable storage medium (e.g., a non-transitory machine-readable storage medium). Although the blocks are illustrated in a particular order, the blocks illustrated in FIG. 5 may be executed in any order and at any time. Also, some of the blocks depicted in method 500 may be omitted without defeating the spirit and scope of this disclosure. The method 500 may be implemented on a network node, such as frames 1 and 2 illustrated in FIGS. 1 and 2, according to any of the examples listed herein.The method 500 begins with a functional ring network in which a ring owner is established and an uplink of one of the FLMs is designated as active uplink in the ring network. In block 502, an FLM in a frame belonging to a group of frames connected in a ring network may generate an uplink discovery packet including information regarding a status and an identifier of a standby uplink port of the FLM. The standby uplink port is configured in a virtual local area network (VLAN) that is unused by other ports of the ring network, such that the standby uplink port does not receive packets from the other ports. At block 504, the FLM may determine that the standby uplink port has a connection to the customer network based on a system description of a link layer discovery protocol (LLDP) packet that the standby uplink port has received from a customer network accessing the group of frames in the ring network.In block 506, the FLM may forward the uplink discovery packet to the standby uplink port via a peripheral component interconnect (PCI) interface of the FLM, the standby uplink port configured to disable the VLAN while forwarding packets received via the PCI interface. In block 508, the FLM may transmit the uplink discovery packet to the customer network via the standby uplink port, wherein the uplink discovery packet is directed to a unicast destination address of an FLM owner in the ring network.In block 510, the owner FLM may monitor receipt of the uplink discovery packet from the customer network via a currently active uplink port in the ring network. In block 512, the owner FLM may determine that the standby uplink port and a set of switches in the customer network between the standby uplink port and the current active uplink port are correctly configured based on the successful reception of the uplink discovery packet.FIG. 6 shows an example of a method 600 for determining whether the uplinks and network switches in the customer network are correctly configured. The method 600 may be performed by FLMs belonging to a group of frames connected in one of multiple ring networks, such as the ring network represented by the first management VLAN 435 and the ring network represented by the second management VLAN 436 in FIG. 4. In some examples, steps 602- 608 may be performed by an FLM, such as FLM 1 to FLM 4, each having an uplink (active or idle), and steps 510- 512 may be performed by an owner FLM, such as owner FLM 2 of FIG. 4. method 600 may be performed as one or more instructions on a machine (e.g., by at least one processor), the one or more instructions being included on at least one machine-readable storage medium (e.g., a non-transitory machine-readable storage medium). Although the blocks are illustrated in a particular order, the blocks illustrated in FIG. 6 may be executed in any order and at any time. Also, some of the blocks depicted in method 600 may be omitted without defeating the spirit and scope of this disclosure. The method 600 may be implemented on a network node, such as frames 1 and 2 illustrated in FIGS. 1 and 2, according to any of the examples described herein.The method 600 begins with a functional ring network having a ring owner established in each ring of the plurality of rings and an uplink identified as an active uplink for each ring network. In block 602, an FLM in a frame belonging to a group of frames connected in one of a plurality of ring networks may generate an uplink discovery packet including information about a status and an identifier of an uplink (active or standby) of the FLM. At block 604, the FLM may determine that the uplink port has a connection to the customer network based on a system description of a link layer discovery protocol (LLDP) packet that the uplink port has received from a customer network accessing resources in the plurality of ring networks.In block 606, the FLM may forward the uplink discovery packet to the uplink port via a peripheral component interconnect (PCI) interface of the FLM. In block 608, the FLM may transmit the uplink discovery packet via the uplink port to the customer network where the uplink discovery packet is directed to a broadcast destination address.In block 610, the owner FLM may monitor receipt of the uplink discovery packet from the customer network via a current active uplink port in the ring network. In block 612, the owner FLM may determine that the standby uplink port and a set of switches in the customer network between the standby uplink port and the current active uplink port are correctly configured based on the successful reception of the uplink discovery packet.FIG. 7 is an example of a computing device 700 having a hardware processor 701 and accessible machine readable instructions stored on a machine readable medium 702 to implement an example of a system for determining uplink connectivity according to one or more disclosed example implementations. In an example, computing device 700 may be a network node, such as an FLM, connected to other network nodes in a ring network. The network node may include an uplink port for sending and receiving packets in a customer network. FIG. 7 shows, by way of example, computing device 700 configured to perform the flow of method 500. However, computing device 700 may also be configured to perform the flow of other methods, techniques, functions, or processes described in this disclosure. In this example of FIG. 7, the machine readable storage medium 702 includes instructions that cause the hardware processor 701 to execute the blocks 502- 512 described above with reference to FIG. 5.A processing element such as processor 701 may include one or more hardware processors, where each hardware processor may have one or more processor cores. In one embodiment, processor 701 may include at least one shared cache that stores data (e.g., computing instructions) used by one or more other components of processor 701. The shared cache may be, for example, locally cached data stored in a memory to allow the components of the processing elements that make up the processor 701 to access it more quickly. In one or more embodiments, the shared cache may include one or more mid-level caches, such as layer 2 (L2), layer 3 (L3), layer 4 (L4), or other cache layers, a last-level cache (LLC), or combinations thereof. Examples of processors include a central processing unit (CPU), a microprocessor. Although not shown in FIG. 5, the processing elements constituting the processor 701 may also include one or more other types of hardware processing components, such as graphics processing units (GPUs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), and / or digital signal processors (DSPs).The processor 701 may be operatively and communicatively coupled to a memory. The memory may be a non-transitory computer readable medium, such as machine readable storage medium 702, configured to store various types of data. For example, the memory may include one or more storage devices including a non-volatile storage device and / or a volatile memory. Volatile memory, such as random access memory (RAM), may be any suitable non-permanent storage device. The non-volatile storage devices may include one or more hard disk drives, optical drives, solid state drives (SSDs), tap drives, flash memory, read only memory (ROM), and / or any other type of memory configured to store data for a period of time after a power failure or shutdown. In certain cases, the non-volatile storage devices may be used to store overflow data when the allocated memory is not large enough to store all working data. The nonvolatile memory devices may also be used to store programs that are loaded into the RAM when such programs are selected for execution.The machine readable storage medium 702 of FIG. 7 may include both volatile and non-volatile, removable, and non-removable media, and may be any electronic, magnetic, optical, or other physical storage device that includes or stores executable instructions, data structures, program modules, or other data accessible by a processor, e.g., firmware, erasable programmable read-only memory (EPROM), random access memory (RAM), non-volatile random access memory (NVRAM), optical disk, solid state drive (SSD), flash memory chips, and the like. The machine readable storage medium may be a non-transitory storage medium, wherein the term "non-transitory" does not include transitory transmission signals.In this specification and claims, certain terms are used that refer to particular system components. As those skilled in the art will appreciate, different parties may refer to a component with different names. In this document, it is not intended to distinguish between components which differ in name but not in function. In this disclosure and in the claims, the terms "including" and "comprising" are used in an open-ended fashion and should therefore be interpreted to mean "including, but not limited to... ". Also, the term "couple" or "couple" is intended to mean either an indirect or direct wired or wireless connection. Thus, if a first device is coupled to a second device, this connection may be through a direct connection or through an indirect connection through other devices and connections. The phrase "based on" is intended to mean "based at least in part on.". Thus, if X is based on Y, X may be a function of Y and any number of other factors.The above is intended to illustrate the principles and various implementations of the present disclosure. Numerous variations and modifications will be apparent to those skilled in the art once the above disclosure is fully understood. The following claims are to be interpreted as including all such variations and modifications.

Claims

A method comprising: generating, by a frame link module (FLM), in a frame belonging to a group of frames connected in a ring network, an uplink discovery packet comprising information regarding a status and an identifier of a standby uplink port of the FLM (502), wherein the standby uplink port is configured in a virtual local area network (VLAN) unused by other ports of the ring network, such that the standby uplink port does not receive packets from the other ports; determining, by the FLM and based on a system description of a link layer discovery protocol (LLDP) packet received by the standby uplink port from a customer network accessing the group of frames in the ring network, that the standby uplink port has a connection to the customer network (504); forwarding, by the FLM, the uplink discovery packet to the standby uplink port via a peripheral component interconnect (PCI) interface of the FLM, wherein the standby uplink port is configured to override the VLAN while forwarding (506) packets received via the PCI interface; sending the uplink discovery packet by the FLM to the customer network via the standby uplink port, the uplink discovery packet directed to a unicast destination address of an owner FLM in the ring network (508); monitoring receipt of the uplink discovery packet from the customer network by the owner FLM via a current active uplink port in the ring network (510); and determining, by the owner FLM and based on successful receipt of the uplink discovery packet, that the standby uplink port and a set of switches in the customer network between the standby uplink port and the current active uplink port are correctly configured (512).The method of claim 1, further comprising: storing the information regarding the status and identifier of the standby uplink port in a set of potential uplink ports; and selecting a new active uplink port from the set of potential uplink ports in response to an error in the current active uplink port.The method of claim 1, further comprising: determining, based on the non-reception of the uplink discovery packet, that the standby uplink port or at least one of the set of switches is misconfigured.The method of claim 1, further comprising removing the information regarding the status and the identifier of the standby uplink port from the set of potential uplink ports in response to expiration of a threshold time from receipt of the uplink discovery packet.The method of claim 1, wherein the uplink discovery packet includes a MAC address of the owner FLM and / or interface information of a switch of the set of switches to which the standby uplink port is connected.The method of claim 1, wherein the uplink discovery packet has a Ring Auto Protection Switching (R-APS) Protocol Data Unit (PDU) format.The method of claim 1, wherein the group of frames constitutes a scalable removable infrastructure (CI) computing resource, and wherein the client network is external to the scalable CI computing resource.The method of claim 1, wherein the owner FLM is the Ring Protection Layer (RPL) owner node according to the International Telecommunication Union (ITU) standard protocol G.8032.The method of claim 1, wherein the group of frames connected in the ring network is configured in a management VLAN to provide an isolated communication path for management traffic between the group of frames.A method comprising: generating, by a frame link module (FLM), in a frame belonging to a group of frames connected in one of a plurality of ring networks, an uplink discovery packet comprising information regarding a status and an identifier of an uplink port of the FLM (602); determining, by the FLM and based on a system description of an LLDP packet received by the uplink port from a customer network accessing resources in the plurality of ring networks, that the uplink port has a connection to the customer network (604); Forwarding, by the FLM, the uplink discovery packet to the uplink port via a peripheral component interconnect (PCI) interface (606) coupled to the uplink port; sending, by the FLM, the uplink discovery packet to a customer network via the uplink port, wherein the uplink discovery packet is directed to a broadcast destination address (608); monitoring, by an owner FLM in a ring network of the plurality (610), receipt of the uplink discovery packet from the customer network via a current active uplink port of the ring network; and determining, by the owner FLM and based on the successful reception of the uplink discovery packet, that the uplink port and a set of switches in the customer network between the uplink port and the current active uplink port are correctly configured ( 612).The method of claim 10, further comprising: determining, based on the non-reception of the uplink discovery packet, that the uplink port or at least one of the group of switches is misconfigured.The method of claim 10, wherein the uplink discovery packet includes an interface information of a switch of the set of switches to which the uplink port is connected.The method of claim 10, wherein the uplink discovery packet has a Ring Auto Protection Switching (R-APS) Protocol Data Unit (PDU) format.The method of claim 10, wherein the FLM and the owner FLM are located in different ring networks.The method of claim 10, wherein the broadcast destination address is a broadcast MAC address.A non-transitory computer readable medium (702) including instructions that, when executed by a processor, cause a frame link module (FLM) in a frame belonging to a group of frames connected in a ring network to: generate an uplink discovery packet including information on a status and an identifier of a standby uplink port of the FLM (502), the standby uplink port configured in a virtual local area network (VLAN) unused by other ports of the ring network such that the standby uplink port does not receive packets from the other ports; based on a system description of a link layer discovery protocol (LLDP) packet received by the standby uplink port from a customer network accessing the group of frames in the ring network, determining that the standby uplink port has a connection to the customer network (504); forwarding the uplink discovery packet to the standby uplink port via a peripheral component interconnect (PCI) interface of the FLM, wherein the standby uplink port is configured to override the VLAN while forwarding (506) packets received via the PCI interface; sending the uplink discovery packet to the customer network via the standby uplink port, the uplink discovery packet directed to a unicast destination address of an owner FLM in the ring network (508); monitoring receipt of the uplink discovery packet from the customer network via a current active uplink port in the ring network (510); and based on successful receipt of the uplink discovery packet, determining that the standby uplink port and a set of switches in the customer network are correctly configured between the standby uplink port and the current active uplink port (512).The non-transitory computer readable medium of claim 16, further comprising instructions that, when executed by the processor, cause the FLM to: store the information regarding the status and identifier of the standby uplink port in a set of potential uplink ports; and select a new active uplink port from the set of potential uplink ports in response to a failure of the current active uplink port.The non-transitory computer readable medium of claim 16, further comprising instructions that, when executed by the processor, cause the FLM to: determine that the standby uplink port or at least one of the set of switches is improperly configured based on the non-receipt of the uplink discovery packet.The non-transitory computer readable medium of claim 16, further comprising instructions that, when executed by the processor, cause the FLM to remove the information regarding the status and identifier of the standby uplink port from the set of potential uplink ports in response to expiration of a threshold time from receipt of the uplink discovery packet.The non-transitory computer readable medium of claim 16, wherein the uplink discovery packet includes at least one MAC address of the owner FLM and / or interface information of a switch of the set of switches to which the standby uplink port is connected.

Images