Management control method of reinforced multi-unit server based on LRM architecture

Abstract

The invention discloses a management control method of a reinforced multi-unit server based on an LRM architecture, and relates to the technical field of server management. Comprising the following steps: adjusting and distributing multi-level dynamic alarm resources based on an LRM architecture: deploying a multi-unit server in the LRM architecture, wherein the multi-unit server comprises a plurality of service blade units, deploying a small alarm processing system on each service blade unit, interconnecting with a complete machine backboard through the small alarm systems, deploying a multi-level alarm system through the complete machine backboard, the multi-level alarm system monitors data in each service blade in real time, summarizes information reported by each small alarm processing system, judges whether each service blade unit has a problem or not according to the monitored data and the reported information, if so, automatically judges the problem alarm level according to the severity of the influence on the running state of the whole machine, and if not, automatically judges the problem alarm level according to the severity of the influence on the running state of the whole machine. And judging whether resource allocation needs to be adjusted according to the problem alarm level.

Description

Technical Field

[0001] The invention discloses a management control method for a reinforced multi-unit server based on an LRM architecture, and relates to the technical field of server management. Background Art

[0002] LRM is a relatively independent unit in terms of system installation structure and function. Fault location can reach LRM level 1, and the fault can be eliminated by replacing LRM. In recent years, LRM architecture has developed rapidly with its advantage of good resistance to harsh environments and has become one of the mainstream architecture forms for reinforced servers.

[0003] The demand for localization of servers in my country is becoming increasingly urgent. In addition, due to the complex working environment of servers, in addition to the difficulty in detecting the status during work, if a system problem occurs, personnel need to arrive at the site, wasting manpower and time, and further improvement is needed. Summary of the invention

[0004] In view of the problems of the prior art, the present invention provides a management and control method for a reinforced multi-unit server based on an LRM architecture, which can dynamically adjust resource allocation according to actual needs to achieve optimal resource utilization and load balancing, thereby improving the performance and reliability of the server and reducing the waste of hardware resources.

[0005] The specific scheme proposed by the present invention is:

[0006] The present invention provides a management and control method for a reinforced multi-unit server based on an LRM architecture, comprising: adjusting and allocating multi-level dynamic alarm resources based on the LRM architecture:

[0007] Deploy multi-unit servers in the LRM architecture. The multi-unit servers contain multiple service blade units. A small alarm processing system is deployed on each service blade unit, which is interconnected with the whole machine backplane through the small alarm system.

[0008] A multi-level alarm system is deployed through the backplane of the whole machine. The data in each service blade is monitored in real time through the multi-level alarm system, and the information reported by each small alarm processing system is summarized. It is judged whether there is a problem with each current service blade unit based on the monitored data and reported information. If so, the problem alarm level is automatically determined according to the severity of the impact on the operating status of the whole machine, and whether resource allocation needs to be adjusted is determined based on the problem alarm level.

[0009] Furthermore, the deployment of a small alarm processing system in the management and control method of a reinforced multi-unit server based on an LRM architecture includes:

[0010] Deploy the embedded chipset and embedded software of the small alarm processing system, and collect and detect the basic hardware data of the mainboard of the service blade unit through the embedded chipset. The basic hardware data includes the voltage, current and temperature of the mainboard; and use the embedded software to detect the system load utilization rate and memory usage of the service blade unit in real time.

[0011] Furthermore, the deployment of a multi-level alarm system in the management and control method of a reinforced multi-unit server based on an LRM architecture includes: deploying an embedded chipset and embedded software of the multi-level alarm system, collecting basic hardware data of the entire machine through the embedded chipset and monitoring the data in each service blade in real time; aggregating information reported by the small alarm processing system of each service blade unit in real time through the embedded software, and judging whether there is a problem with each current service blade unit based on the monitored data and the reported information.

[0012] Furthermore, the management and control method of a reinforced multi-unit server based on the LRM architecture described herein automatically determines the problem alarm level according to the severity of the impact on the operating status of the entire machine, including: dividing the levels into four types, namely dangerous, important, general, and prompt, wherein the dangerous level indicates that the problem is a serious fault and an unrecoverable problem occurs in the server; the important level indicates that the problem is relatively serious and affects the normal use of the server; the general level indicates that the problem is not serious and may affect the normal use of the server; the prompt level indicates a potential problem that currently exists and may affect the normal use of the server.

[0013] Furthermore, the method for managing and controlling a reinforced multi-unit server based on an LRM architecture, judging whether resource allocation needs to be adjusted according to the problem alarm level, includes:

[0014] If there is a dangerous level, report it immediately;

[0015] If an important or general level alarm occurs, the multi-level alarm system will determine the source of the problem based on the information reported by each service blade unit, clean up zombie resources, and dynamically adjust the priority to adjust the process or thread to other service blade units with idle loads until the alarm is eliminated;

[0016] If there is a warning level, abnormal indications are sent regularly through the small alarm system, waiting for processing by the multi-level alarm system.

[0017] The present invention also provides a management control device for a reinforced multi-unit server based on an LRM architecture, which adjusts and allocates multi-level dynamic alarm resources based on the LRM architecture, including a deployment module, a small alarm system and a multi-level alarm system.

[0018] The deployment module deploys a multi-unit server in the LRM architecture. The multi-unit server contains multiple service blade units. A small alarm processing system is deployed on each service blade unit, which is interconnected with the whole machine backplane through the small alarm system.

[0019] The deployment module deploys a multi-level alarm system through the backplane of the whole machine, monitors the data in each service blade in real time through the multi-level alarm system, and summarizes the information reported by each small alarm processing system. It determines whether there is a problem with each current service blade unit based on the monitored data and reported information. If so, it automatically determines the problem alarm level based on the severity of the impact on the operating status of the whole machine, and determines whether resource allocation needs to be adjusted based on the problem alarm level.

[0020] Furthermore, the deployment module of the management control device for reinforcing a multi-unit server based on the LRM architecture deploys a small alarm processing system, including:

[0021] Deploy the embedded chipset and embedded software of the small alarm processing system, and collect and detect the basic hardware data of the service blade unit motherboard through the embedded chipset. The basic hardware data includes the motherboard voltage, current, and temperature; use the embedded software to detect the system load utilization rate and memory usage of the service blade unit in real time.

[0022] Furthermore, the deployment module of the management control device of the reinforced multi-unit server based on the LRM architecture deploys a multi-level alarm system, including: deploying an embedded chipset and embedded software of the multi-level alarm system, collecting basic hardware data of the whole machine through the embedded chipset and monitoring the data in each service blade in real time; aggregating information reported by the small alarm processing system of each service blade unit in real time through the embedded software, and judging whether there is a problem with each current service blade unit according to the monitored data and the reported information.

[0023] Furthermore, the multi-level alarm system of the management and control device of the reinforced multi-unit server based on the LRM architecture automatically determines the problem alarm level according to the severity of the impact on the operating status of the whole machine, including: dividing the levels into four types, namely dangerous, important, general, and prompt. The dangerous level indicates that the problem is a serious fault and the server has an unrecoverable problem; the important level indicates that the problem is relatively serious and affects the normal use of the server; the general level indicates that the problem is not serious and may affect the normal use of the server; the prompt level indicates a potential problem that currently exists and may affect the normal use of the server.

[0024] Furthermore, the multi-level alarm system of the management control device for reinforcing multi-unit servers based on the LRM architecture determines whether resource allocation needs to be adjusted according to the problem alarm level, including:

[0025] If there is a dangerous level, report it immediately;

[0026] If an important or general alarm occurs, the multi-level alarm system determines the source of the problem based on the information reported by each service blade unit, cleans up zombie resources, and dynamically adjusts the priority to move the process or thread to other service blade units with idle loads until the alarm is eliminated;

[0027] If there is a warning level, the small alarm system will periodically send abnormal indications, waiting for the multi-level alarm system to process.

[0028] The benefits of the present invention are:

[0029] The method of the present invention can dynamically adjust resource allocation according to actual needs to achieve optimal resource utilization and load balancing. While improving the performance and reliability of the server, it can also reduce the waste of hardware resources. At the same time, it can locate fault problems, play a role in remote status detection, and can also remotely troubleshoot problems, saving manpower and time, and has good promotion and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of the application framework of the device of the present invention.

[0031] Figure 2 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION

[0032] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0033] Example 1

[0034] The present invention provides a management and control method for a reinforced multi-unit server based on an LRM architecture, comprising: adjusting and allocating multi-level dynamic alarm resources based on the LRM architecture:

[0035] Step 1: Deploy a multi-unit server in the LRM architecture. The multi-unit server includes multiple service blade units. A small alarm processing system is deployed on each service blade unit, and the small alarm system is interconnected with the whole machine backplane.

[0036] The deployment of the small alarm processing system includes:

[0037] Deploy the embedded chipset and embedded software of the small alarm processing system, and collect and detect the basic hardware data of the mainboard of the service blade unit through the embedded chipset. The basic hardware data includes the voltage, current and temperature of the mainboard; and use the embedded software to detect the system load utilization rate and memory usage of the service blade unit in real time.

[0038] Step 2: Deploy a multi-level alarm system through the backplane of the whole machine, monitor the data in each service blade in real time through the multi-level alarm system, and summarize the information reported by each small alarm processing system. Determine whether there is a problem with each current service blade unit based on the monitored data and reported information. If so, automatically determine the problem alarm level based on the severity of the impact on the operating status of the whole machine, and determine whether resource allocation needs to be adjusted based on the problem alarm level.

[0039] The deployment of the multi-level alarm system includes: deploying an embedded chipset and embedded software of the multi-level alarm system, collecting basic hardware data of the entire machine through the embedded chipset and monitoring the data in each service blade in real time; aggregating information reported by the small alarm processing system of each service blade unit in real time through the embedded software, and judging whether there is a problem with each current service blade unit based on the monitored data and reported information.

[0040] The problem alarm level is automatically determined according to the severity of the impact on the operating status of the entire machine, and may further include: dividing the levels into four types, namely dangerous, important, general, and warning. The dangerous level indicates that the problem is a serious fault and an unrecoverable problem has occurred in the server; the important level indicates that the problem is relatively serious and affects the normal use of the server; the general level indicates that the problem is not serious and may affect the normal use of the server; the warning level indicates a potential problem that currently exists and may affect the normal use of the server.

[0041] The determining whether resource allocation needs to be adjusted according to the problem alarm level includes:

[0042] If there is a dangerous level, report it immediately;

[0043] If an important or general level alarm occurs, the multi-level alarm system will determine the source of the problem based on the information reported by each service blade unit, clean up zombie resources, and dynamically adjust the priority to adjust the process or thread to other service blade units with idle loads until the alarm is eliminated;

[0044] If there is a warning level, abnormal indications are sent regularly through the small alarm system, waiting for processing by the multi-level alarm system.

[0045] For example, when the server problem alarm level is dangerous and an unrecoverable alarm occurs, it is directly reported to the operator. For example, the management system detects that a blade is powered on, but the network is not connected. The management system first adjusts the hardware resources, uses the serial port to detect the blade system status, and tries to restore the network through the serial port. If the recovery fails, it is reported to the operator; if the recovery is successful, the alarm is canceled.

[0046] When a server problem occurs and the alarm level reaches an important level, for example, the processor occupancy rate of a blade in the entire machine reaches more than 80%, if it is not handled, problems such as system freezing will occur. The multi-level alarm system will issue an adjustment command to adjust the process or thread to other core performance with less load, thereby improving the processing capacity of the entire machine until the alarm is eliminated.

[0047] When a server problem occurs and the alarm level is a general level, the resources of the entire machine can be deployed for processing. For example, the management system detects that the temperature of a service blade has risen, and the range is above the normal operating temperature threshold and below the irrecoverable threshold. The multi-level alarm system determines it as a "general problem" and adjusts the fan speed of the blade to reduce the temperature to below the normal operating temperature threshold. If it fails, an automatic frequency reduction command is issued to dynamically adjust the power consumption of the entire machine until the alarm is eliminated. When a "prompt level" alarm appears on the server, the small alarm system regularly sends abnormal indications to the next level and waits for the previous level to process.

[0048] Example 2

[0049] The present invention also provides a management and control device for a reinforced multi-unit server based on an LRM architecture, which adjusts and allocates multi-level dynamic alarm resources based on the LRM architecture, including a deployment module, a small alarm system and a multi-level alarm system.

[0050] The deployment module deploys a multi-unit server in the LRM architecture. The multi-unit server contains multiple service blade units. A small alarm processing system is deployed on each service blade unit, which is interconnected with the whole machine backplane through the small alarm system.

[0051] The deployment module deploys a multi-level alarm system through the backplane of the whole machine, monitors the data in each service blade in real time through the multi-level alarm system, and summarizes the information reported by each small alarm processing system. It determines whether there is a problem with each current service blade unit based on the monitored data and reported information. If so, it automatically determines the problem alarm level based on the severity of the impact on the operating status of the whole machine, and determines whether resource allocation needs to be adjusted based on the problem alarm level.

[0052] As the information interaction and execution process between the modules in the above-mentioned device are based on the same concept as the embodiment of the method of the present invention, the specific contents can be found in the description of the embodiment of the method of the present invention and will not be repeated here.

[0053] Similarly, the device of the present invention can dynamically adjust resource allocation according to actual needs to achieve optimal resource utilization and load balancing. While improving the performance and reliability of the server, it can also reduce the waste of hardware resources. At the same time, it can locate fault problems, play a role in remote status detection, and can also remotely troubleshoot problems, saving manpower and time, and has good promotion and application value.

[0054] It should be noted that not all steps and modules in the above-mentioned processes and device structures are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted as needed. The system structure described in the above-mentioned embodiments can be a physical structure or a logical structure, that is, some modules may be implemented by the same physical entity, or some modules may be implemented by multiple physical entities, or some components in multiple independent devices may be implemented together.

[0055] The above-described embodiments are only preferred embodiments for fully illustrating the present invention, and the protection scope of the present invention is not limited thereto. Equivalent substitutions or changes made by those skilled in the art based on the present invention are within the protection scope of the present invention. The protection scope of the present invention shall be subject to the claims.

Claims

1. A management and control method for a reinforced multi-unit server based on an LRM architecture, characterized in that include: Based on the LRM architecture, multi-level dynamic alarm resource adjustment and allocation are performed: Deploy multi-unit servers in the LRM architecture. The multi-unit servers contain multiple service blade units. A small alarm processing system is deployed on each service blade unit, which is interconnected with the whole machine backplane through the small alarm system. A multi-level alarm system is deployed through the backplane of the whole machine. The data in each service blade is monitored in real time through the multi-level alarm system, and the information reported by each small alarm processing system is summarized. It is judged whether there is a problem with each current service blade unit based on the monitored data and reported information. If so, the problem alarm level is automatically determined according to the severity of the impact on the operating status of the whole machine, and whether resource allocation needs to be adjusted is determined based on the problem alarm level.

2. According to the management and control method of the reinforced multi-unit server based on the LRM architecture according to claim 1, it is characterized by: The deployment of the small alarm processing system includes: Deploy the embedded chipset and embedded software of the small alarm processing system, and collect and detect the basic hardware data of the service blade unit motherboard through the embedded chipset. The basic hardware data includes the motherboard voltage, current, and temperature; use the embedded software to detect the system load utilization rate and memory usage of the service blade unit in real time.

3. According to the management and control method of the reinforced multi-unit server based on the LRM architecture of claim 1, it is characterized by: The deployment of the multi-level alarm system includes: deploying an embedded chipset and embedded software of the multi-level alarm system, collecting basic hardware data of the entire machine through the embedded chipset and monitoring the data in each service blade in real time; aggregating information reported by the small alarm processing system of each service blade unit in real time through the embedded software, and judging whether there is a problem with each current service blade unit based on the monitored data and reported information.

4. According to the management and control method of the reinforced multi-unit server based on the LRM architecture of claim 1, it is characterized by: The problem alarm level is automatically determined according to the severity of the impact on the operating status of the entire machine, including: dividing the levels into four types, namely dangerous, important, general, and prompt. The dangerous level indicates that the problem is a serious fault and an unrecoverable problem occurs in the server; the important level indicates that the problem is relatively serious and affects the normal use of the server; the general level indicates that the problem is not serious and may affect the normal use of the server; the prompt level indicates a potential problem that currently exists and may affect the normal use of the server.

5. According to the management and control method of the reinforced multi-unit server based on the LRM architecture of claim 1, it is characterized by: The determining whether resource allocation needs to be adjusted according to the problem alarm level includes: If there is a dangerous level, report it immediately; If an important or general level alarm occurs, the multi-level alarm system will determine the source of the problem based on the information reported by each service blade unit, clean up zombie resources, and dynamically adjust the priority to adjust the process or thread to other service blade units with idle loads until the alarm is eliminated; If there is a warning level, abnormal indications are sent regularly through the small alarm system, waiting for processing by the multi-level alarm system.

6. A management and control device for a reinforced multi-unit server based on an LRM architecture, which adjusts and allocates multi-level dynamic alarm resources based on the LRM architecture, and is characterized by: Including deployment modules, small alarm system and multi-level alarm system, The deployment module deploys a multi-unit server in the LRM architecture. The multi-unit server contains multiple service blade units. A small alarm processing system is deployed on each service blade unit, which is interconnected with the whole machine backplane through the small alarm system. The deployment module deploys a multi-level alarm system through the backplane of the whole machine, monitors the data in each service blade in real time through the multi-level alarm system, and summarizes the information reported by each small alarm processing system. It determines whether there is a problem with each current service blade unit based on the monitored data and reported information. If so, it automatically determines the problem alarm level based on the severity of the impact on the operating status of the whole machine, and determines whether resource allocation needs to be adjusted based on the problem alarm level.

7. A management and control device for a reinforced multi-unit server based on an LRM architecture according to claim 6, characterized in that The deployment module deploys a minor alarm processing system, including: Deploy the embedded chipset and embedded software of the small alarm processing system, and collect and detect the basic hardware data of the service blade unit motherboard through the embedded chipset. The basic hardware data includes the motherboard voltage, current, and temperature; use the embedded software to detect the system load utilization rate and memory usage of the service blade unit in real time.

8. The management control device for a reinforced multi-unit server based on an LRM architecture according to claim 6, characterized in that The deployment module deploys a multi-level alarm system, including: deploying an embedded chipset and embedded software of the multi-level alarm system, collecting basic hardware data of the entire machine through the embedded chipset and monitoring the data in each service blade in real time; aggregating information reported by the small alarm processing system of each service blade unit in real time through the embedded software, and judging whether there is a problem with each current service blade unit based on the monitored data and reported information.

9. The management control device for a reinforced multi-unit server based on an LRM architecture according to claim 6, characterized in that The multi-level alarm system automatically determines the problem alarm level according to the severity of the impact on the operating status of the entire machine, including: dividing the levels into four types, namely dangerous, important, general, and prompt. The dangerous level indicates that the problem is a serious fault and an unrecoverable problem occurs on the server; the important level indicates that the problem is relatively serious and affects the normal use of the server; the general level indicates that the problem is not serious and may affect the normal use of the server; the prompt level indicates a potential problem that currently exists and may affect the normal use of the server.

10. The management control device for a reinforced multi-unit server based on an LRM architecture according to claim 6, characterized in that The multi-level alarm system determines whether resource allocation needs to be adjusted according to the problem alarm level, including: If there is a dangerous level, report it immediately; If an important or general alarm occurs, the multi-level alarm system determines the source of the problem based on the information reported by each service blade unit, cleans up zombie resources, and dynamically adjusts the priority to move the process or thread to other service blade units with idle loads until the alarm is eliminated; If there is a warning level, the small alarm system will periodically send abnormal indications and wait for the multi-level alarm system to process.

Images