A multi-node server management system and a multi-node server

Abstract

The application discloses a kind of multi-node server's management system and multi-node server, applied to server management technical field, including respectively with power, fan and multiple BMC connection selection module, with the processor of selection module connection and respectively set in each node of multi-node server in BMC, processor will be through selection module according to preset priority round robin control each BMC and power and fan connection.Visibly, in this scheme, the management right of CMC is decentralized to each BMC, and each BMC is used to monitor and manage power and fan, the management architecture is simple, the stability of the management system is enhanced, and the use scenario is expanded.In addition, CMC is omitted, and the management cost is reduced.

Description

Technical Field

[0001] This invention relates to the field of server management technology, and in particular to a management system for a multi-node server and a multi-node server. Background Technology

[0002] As users' computing demands increase, multi-node servers have emerged. In existing technologies, to centrally manage power and fans in multi-node servers, a two-tier management architecture, such as the Chassis Management Controller (CMC), has been introduced. This CMC two-tier architecture includes the CMC and multiple Chassis Management Controllers (BMCs) on each node of the multi-node server. However, the CMC two-tier management architecture suffers from drawbacks such as architectural complexity and complex customer data center management networks, resulting in very limited use of CMCs at customer sites, and the CMC itself is also costly. Summary of the Invention

[0003] The purpose of this invention is to provide a management system and a multi-node server for multi-node servers. The management architecture is simple, the stability of the management system is enhanced, the application scenarios are expanded, and the CMC is omitted, thus reducing management costs.

[0004] To address the aforementioned technical problems, this invention provides a management system for a multi-node server. The multi-node server includes a power supply and a fan. The management system for the multi-node server includes:

[0005] Each BMC is set in each node of the multi-node server, and each BMC is used to monitor the power supply and fan connected to it.

[0006] Selection modules are respectively connected to the power supply, the fan, and the plurality of BMCs;

[0007] The processor connected to the selection module is used to control the connection of each BMC to the power supply and the fan in a round-robin manner according to a preset priority through the selection module.

[0008] Preferably, each BMC is specifically used to acquire power data of the power supply connected to it and fan data of the fan, and to control the fan speed according to the temperature data of the multi-node server.

[0009] Preferably, each of the BMCs is also used to display the power data and / or fan data through a display module when a power data query command and / or fan data query command is received.

[0010] Preferably, the processor is further configured to, upon receiving a BMC control command, control the BMC corresponding to the BMC control command to connect with the power supply and the fan.

[0011] Preferably, the processor is also connected to a plurality of the BMCs; the processor is further configured to send a notification instruction to the BMCs connected to the power supply and the fan after controlling the BMCs to connect to the power supply and the fan, so that the BMCs can monitor the power supply and the fan connected to them after receiving the notification instruction.

[0012] Preferably, the selection module is an I2C MUX.

[0013] Preferably, the processor is also connected to multiple BMCs; the selection module controls the connection of each BMC to the power supply and the fan in a round-robin fashion according to a preset priority, including:

[0014] Upon receiving a monitoring completion message from the BMC currently monitoring the power supply and the fan, the selection module uses a preset priority cycle to control the next BMC to connect to the power supply and the fan.

[0015] Preferably, each BMC is connected to the processor via a GPIO interface.

[0016] Preferably, the selection module controls the connection of each BMC to the power supply and the fan in a round-robin fashion according to a preset priority, including:

[0017] When the current BMC is connected to the power supply and the fan, a timer starts. When the timer reaches a preset duration, the selection module controls the next BMC to connect to the power supply and the fan in a preset priority cycle.

[0018] To address the aforementioned technical problems, the present invention also provides a multi-node server, including a power supply and a fan, and a management system for the multi-node server as described above, wherein the management system is connected to the power supply and the fan respectively.

[0019] This invention provides a management system and a multi-node server for a multi-node server, applied in the field of server management technology. It includes a selection module connected to a power supply, fans, and multiple BMCs (Block Controllers), a processor connected to the selection module, and BMCs located in each node of the multi-node server. The processor, through the selection module, controls the connection of each BMC to the power supply and fans in a round-robin fashion according to a preset priority. Therefore, this solution delegates the management authority of the CMC (Controlled Control Center) to each BMC, allowing each BMC to monitor and manage the power supply and fans. This simplifies the management architecture, enhances the stability of the management system, expands its application scenarios, and, by omitting the CMC, reduces management costs. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the prior art and embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a schematic diagram of the structure of a multi-node server management system provided by the present invention. Detailed Implementation

[0022] The core of this invention is to provide a management system and a multi-node server for multi-node servers. The management architecture is simple, the stability of the management system is enhanced, the application scenarios are expanded, and the CMC is omitted, thus reducing management costs.

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please refer to Figure 1 , Figure 1 This is a schematic diagram of the structure of a multi-node server management system provided by the present invention.

[0025] A multi-node server includes a power supply and fans, and its management system includes:

[0026] Each BMC 1 (Baseboard Management Controller) is located in each node of the multi-node server. Each BMC 1 is used to monitor the power supply and fans connected to it.

[0027] Selection module 2 is connected to the power supply, fan, and multiple BMCs 1 respectively;

[0028] The processor 3, connected to the selection module 2, is used to control the connection of each BMC 1 to the power supply and fan in a round-robin manner according to a preset priority through the selection module 2.

[0029] The multi-node server can be a 4-node server or an 8-node server, etc., and this application does not make any special limitation.

[0030] To simplify the management architecture of multi-node servers, the management system of the multi-node server in this application includes a selection module 2, a processor 3, and BMC 1 set in each node of the multi-node server. In addition, the management authority of CMC is delegated to each BMC 1. In this way, the power supply and fans can be monitored through the BMC 1 in each node, which makes it convenient for users to query the monitoring data of power supply and fans through the BMC 1.

[0031] Specifically, selection module 2 connects to each BMC 1, as well as to the power supply, fans, and processor 3. Furthermore, priorities can be pre-set for each BMC 1. Taking a 4-node server as an example, the management system includes four BMC 1s, with priorities set for BMC 1 in node 1, node 2, node 3, and node 4. Processor 3 controls selection module 2 to control the connection of each BMC 1 to the power supply and fans in a round-robin fashion according to the preset priorities. When connected to the power supply and fans, BMC 1 monitors them, such as acquiring power data and fan data, and can also control fan speed based on the current temperature in the multi-node server.

[0032] In this application, the processor 3 may be, but is not limited to, a CPLD (Complex Programmable Logic Device), and the power supply may be a PSU (Power Supply Unit). A multi-node server may have one or more fans, and the number of power supplies may be, but is not limited to, one.

[0033] As can be seen, this solution simplifies the multi-level CMC architecture by delegating the management authority of the CMC to each BMC 1, which then monitors and manages the power supply and fans. This simplifies the management architecture, enhances the stability of the management system, expands the application scenarios, and reduces management costs by omitting the CMC.

[0034] Based on the above embodiments:

[0035] In a preferred embodiment, each BMC 1 is specifically used to acquire power data of the power supply connected to it and fan data of the fan, and to control the fan speed according to the temperature data of the multi-node server.

[0036] Specifically, when connected to the power supply and fan, BMC 1 monitors the power supply and fan, such as acquiring power data and fan data, and also acquiring temperature data collected by temperature sensors in the multi-node server, and controlling the fan speed based on the current temperature data in the multi-node server.

[0037] In addition, the power data here may include one or more combinations of power supply input voltage, output voltage, fault status, etc., and the fan data may include one or more combinations of fan speed, fault information, etc.

[0038] As can be seen, this solution simplifies the CMC multi-level architecture, but still retains centralized management of power supplies and fans in terms of management functions, reducing complexity and lowering costs.

[0039] In a preferred embodiment, each BMC 1 is also used to display power data and / or fan data through a display module when a power data query command and / or fan data query command is received.

[0040] Specifically, in this application, each BMC 1 located in each node of the multi-node server can monitor the power supply and fan. When a user wants to query the power supply data and / or fan data, they can send a power supply data query command and / or fan data query command to the BMC 1. When the BMC 1 receives the power supply data query command and / or fan data query command, it displays the power supply data and / or fan data through the display module. It can be seen that this solution makes it convenient for users to query the power supply and fan data through the nodes they manage.

[0041] In a preferred embodiment, the processor 3 is also configured to, upon receiving a control command for BMC 1, control the connection between BMC 1 corresponding to the control command for BMC 1 and the power supply and fan.

[0042] In this application, the processor 3 can automatically control each BMC 1 to connect to the power supply and fan in a round-robin fashion according to a preset priority via the selection module 2, thereby enabling the BMC 1 to monitor the power supply and fan. In addition, the management system can also control a specified BMC 1 to monitor the power supply and fan according to user instructions. Specifically, when the processor 3 receives a BMC 1 control instruction from the user, it controls the BMC 1 corresponding to the BMC 1 control instruction to connect to the power supply and fan, so that the BMC 1 can monitor the power supply and fan. In practical applications, after the BMC 1 has finished monitoring the power supply and fan, the processor 3 can continue to control it in a round-robin fashion starting from the next BMC 1 according to a preset priority.

[0043] As can be seen, in addition to automatically controlling BMC 1 to monitor the power supply and fans, the management system can also control a designated BMC 1 to monitor the power supply and fans according to user instructions, making the control method flexible.

[0044] In a preferred embodiment, the processor 3 is also connected to a plurality of BMCs 1; the processor 3 is also configured to send a notification instruction to the BMCs 1 connected to the power supply and fan after controlling the BMCs 1 to connect to the power supply and fan, so that the BMCs 1 can monitor the power supply and fan connected to them after receiving the notification instruction.

[0045] Considering that processor 3 will control each BMC 1 to monitor the power supply and fan in a round-robin fashion according to a preset priority, in order to improve control reliability, in this application, after controlling a BMC 1 to connect to the power supply and fan, processor 3 will send a notification command to that BMC 1, so that the BMC 1 can monitor the power supply and fan connected to it after receiving the notification command. Specifically, the notification command can be an interrupt signal. Therefore, this control method can reliably and stably enable each BMC 1 to monitor the power supply and fan.

[0046] In a preferred embodiment, module 2 is selected as an I2C MUX.

[0047] In this application, module 2 can be, but is not limited to, an I2C MUX, which has the advantages of simple structure, low power consumption, and strong anti-interference.

[0048] In a preferred embodiment, the processor 3 is also connected to multiple BMCs 1; the selection module 2 controls the connection of each BMC 1 to the power supply and fan in a round-robin manner according to a preset priority, including:

[0049] Upon receiving a monitoring completion message from BMC 1, which is currently monitoring the power supply and fan, the selection module 2 controls the next BMC 1 to connect to the power supply and fan in a round-robin fashion according to a preset priority.

[0050] To ensure that each BMC 1 can fully monitor the power supply and fan, in this application, after BMC 1 completes the monitoring of the power supply and fan, it will send a monitoring completion message to the processor 3. When the processor 3 receives the monitoring completion message sent by BMC 1, it will use the selection module 2 to control the next BMC 1 to connect to the power supply and fan in a round-robin manner according to the preset priority.

[0051] In practical applications, when monitoring power supplies and fans, BMC 1 can send power data acquisition commands to the power supply and fan data acquisition commands to the fans. After acquiring the power data sent by the power supply and the fan data sent by the fans, and controlling the fan speed based on the temperature of the multi-node server during this process, the monitoring can be considered complete.

[0052] In a preferred embodiment, each BMC 1 is connected to the processor 3 via a GPIO interface.

[0053] In this application, each BMC 1 is connected to the processor 3 via a GPIO interface, thereby ensuring the real-time communication between each BMC 1 and the processor 3. Of course, each BMC 1 and the processor 3 can also be connected via other interfaces, which are not specifically limited here.

[0054] In one preferred embodiment, the selection module 2 controls the connection of each BMC 1 to the power supply and fan in a round-robin manner according to a preset priority, including:

[0055] When the current BMC 1 is connected to the power supply and fan, a timer starts. When the timer reaches the preset duration, the next BMC 1 is connected to the power supply and fan in a preset priority cycle by selecting module 2.

[0056] To simplify the control process, in this application, the processor 3 starts timing each time it controls a BMC 1 to connect to the power supply and fan. When the timer reaches a preset duration, it is determined that the BMC 1 has completed monitoring of the power supply and fan. At this point, the processor 3 controls the next BMC 1 to connect to the power supply and fan in a preset priority cycle. The preset duration can be set according to the actual application's requirement for the BMC 1 to monitor the fan and power supply once.

[0057] It is evident that this method enables each BMC 1 to perform cyclic control of the fan and power supply, and the control process is simple and reliable.

[0058] The present invention also provides a multi-node server, including a power supply and a fan, and a management system for the multi-node server as described above, wherein the management system for the multi-node server is connected to the power supply and the fan respectively.

[0059] For a description of the management system in the multi-node server provided by this invention, please refer to the above-described management system embodiments. This application does not impose any special limitations here.

[0060] It should be noted that, in this specification, the terms "comprising," "including," or any other variations thereof are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0061] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A management system for a multi-node server, characterized in that, The multi-node server includes a power supply and a fan, and the management system of the multi-node server includes: Each BMC is configured in each node of the multi-node server. Each BMC is used to obtain power data of the power supply and fan data of the fan when connected to the fan and power supply, and to control the fan speed according to the temperature data of the multi-node server. Selection modules are respectively connected to the power supply, the fan, and the plurality of BMCs; The processor connected to the selection module is used to control the connection of each BMC to the power supply and the fan in a round-robin manner according to a preset priority through the selection module; The processor is also connected to a plurality of the BMCs; the processor is also configured to send a notification instruction to the BMCs connected to the power supply and the fan after controlling the BMCs to connect to the power supply and the fan, so that the BMCs monitor the power supply and the fan connected to them after receiving the notification instruction; The processor, through the selection module, controls the connection of each BMC to the power supply and the fan in a round-robin fashion according to a preset priority, including: When the processor receives a monitoring completion message from the BMC currently monitoring the power supply and the fan, it uses the selection module to control the next BMC to connect to the power supply and the fan in a round-robin fashion according to a preset priority. Alternatively, when the current BMC is connected to the power supply and the fan, a timer is started, and when the timer reaches a preset duration, the selection module controls the next BMC to connect to the power supply and the fan in a preset priority cyclical order. The power data includes one or more of the following: power input voltage, output voltage, and fault status; the fan data includes one or more of the following: fan speed and fault information.

2. The management system for a multi-node server as described in claim 1, characterized in that, Each of the BMCs is also used to display the power data and / or fan data through the display module when it receives a power data query command and / or a fan data query command.

3. The management system for a multi-node server as described in claim 1, characterized in that, The processor is also configured to, upon receiving a BMC control command, control the BMC corresponding to the BMC control command to connect with the power supply and the fan.

4. The management system for a multi-node server as described in claim 1, characterized in that, The selected module is I2CMUX.

5. The management system for a multi-node server as described in claim 1, characterized in that, Each BMC is connected to the processor via a GPIO interface.

6. A multi-node server, characterized in that, It includes a power supply and a fan, and also includes a management system for a multi-node server as described in any one of claims 1 to 5, wherein the management system for the multi-node server is connected to the power supply and the fan respectively.

Images