CN107995287A - A method for remotely monitoring the health status of data center nodes through IPMI - Google Patents

Abstract

The present invention provides a method for remotely monitoring the health status of data center nodes through IPMI. Through the IPMI protocol, the running status of servers is monitored, including CPU, hard disk, fan, memory, power supply, motherboard, etc., and all servers can be remotely monitored through the monitoring terminal. It can monitor the server load, CPU and memory usage, network fluctuations, etc. in real time, and judge according to the safety parameter warning line set by the administrator. When the threshold value is exceeded, it will automatically The policies are processed and messages are automatically sent to managers. This method can save a lot of manpower during the daily maintenance of server operation, and avoid downtime or other abnormalities caused by various hidden dangers in the data center. It will greatly reduce the risk of data center node downtime and ensure the reliability of the data center.

Description

A method for remotely monitoring the health status of data center nodes through IPMI

technical field

The invention relates to the field of servers, in particular to a method for remotely monitoring the health status of data center nodes through IPMI.

Background technique

In the current big data generation of cloud computing, the stability of server nodes in data centers plays a decisive role in data security. Generally, in large data centers, servers continue to work 365x24 hours without stopping. During the operation of the current system, it is inevitable There will be various component failures, high-load operation, and high temperature. When these problems occur, it will affect the normal use of the server, and seriously affect the use of the entire system, resulting in downtime or data loss.

Contents of the invention

In order to overcome the deficiencies in the above-mentioned prior art, the present invention provides a method for remotely monitoring the health status of data center nodes through IPMI, including: a monitoring terminal, characterized in that the method includes:

The monitoring terminal and the data center are remotely connected to the same local area network, so that the monitoring terminal communicates with each node of the data center;

The monitoring terminal accesses each node of the data center through the IPMI protocol;

The monitoring terminal obtains temperature information, fan information, voltage information, network card working conditions and operating system working conditions of each node;

The monitoring terminal compares the obtained information with the corresponding threshold value, and when the obtained data information exceeds the threshold value, an alarm message is sent in the data center to prompt the maintenance personnel.

Preferably, the method includes:

The monitoring terminal monitors the temperature information, fan information, voltage information, network card working conditions and operating system working conditions of each node every time a preset time interval passes.

Preferably, the monitoring terminal analyzes the received data information, classifies the analyzed data information through corresponding alarm parameters, forms corresponding data information for data exceeding the threshold, and sends an alarm message in the data center to prompt maintenance personnel , and at the same time, the alarm information is formed into an alarm log;

The monitoring terminal presents the obtained data information and alarm information to the maintenance personnel, and at the same time receives the control instructions from the maintenance personnel to the monitoring terminal;

The monitoring terminal provides a unified CUI interface for maintenance personnel to query and browse historical logs and alarm information and set alarm parameters.

Preferably, the temperature information includes CPU temperature, motherboard temperature, backplane temperature, and hot-swap module temperature;

Voltage information includes CPU voltage, motherboard voltage, SCSI backplane voltage, and hot-swap module voltage

Fan information includes CPU fans, motherboard fans, backplane fans, and hot-swap module fans.

Preferably, the monitoring terminal has an I2C device interface, and each node is provided with an I2C Slave interface;

The monitoring terminal generates a clock through the I2C device interface to initiate communication to the I2C Slave interface of each node, and each node responds through the I2C Slave interface

The I2C device interface of the monitoring terminal and the I2C Slave interface of each node transmit data through the IPMI protocol.

Preferably, the step monitoring terminal accessing each node of the data center through the IPMI protocol also includes:

The monitoring terminal configures the alarm information, including: node name, node device name, node alarm event name, node alarm description, node alarm trigger value;

The monitoring terminal configures the types of alarm information, including: alarm power supply and distribution, alarm environment, alarm security;

The monitoring terminal is configured with alarm masking control, node alarm masking, node device alarm masking, and node alarm event masking, node alarms are masked by configuring masking methods and masking time periods, and masking control is configured;

The monitoring terminal configures the alarm level and configures the node level;

Configure node CPU, hard disk, and motherboard alarms as high-level alarms, configure node temperature alarms as intermediate-level alarms, and configure node software alarms as low-level alarms;

Configure the node power supply and distribution class as high-level alarm, configure the node environment class as medium-level alarm, and configure the node security class as low-level alarm;

According to the alarm rules, when multiple alarms are generated, the monitoring terminal will give priority to the alarm with the highest alarm level according to the configured alarm level;

Or get multiple alarms in the same time period, according to the configured alarm level, give priority to the alarm with higher alarm level;

When multiple alarms are generated, the low-level alarms are cached for a preset time and sent out;

When a node alarms multiple times within the preset time period, it is considered as a high-frequency alarm, and a high-frequency alarm prompt will be issued after the preset time period.

Preferably, the monitoring terminal screens the alarm keywords, combines the same alarm keywords belonging to the same time interval, and counts the frequency of alarm occurrences, counts the frequency of alarm occurrences according to the time interval, and displays them to maintenance personnel.

Preferably, the step monitoring terminal accessing each node of the data center through the IPMI protocol also includes:

The monitoring terminal encapsulates the configured alarm information, alarm information type, and alarm shielding control information, and sends the encapsulated information to each node in the data center

Encapsulation information includes: configure alarm information MAC address layer, configure alarm information data processing terminal IP layer, configure alarm information data frame segment;

In the configuration alarm information data frame segment, when the lowest bit of the frame segment is 0, it indicates the address code of the request/reply response component; when the lowest bit is 1, it indicates the software ID of the request/reply response; the upper 7 bits of the byte Indicates the specific address code and software ID;

The parity information code is set in the configuration alarm information data frame segment. When the parity information code is an even number, it means that the message is a request. When the parity information code is an odd number, it means that the message is a response. When the message is a basic control request Or the byte is 00h and 01h in status response;

There is also a serial number generated by the requester in the configuration alarm information data frame segment, which is used to identify different requests when the monitoring terminal needs to send multiple requests; the serial number is the subcomponent number of the component that responds/receives the message or subaddress number.

As can be seen from the above technical solutions, the present invention has the following advantages:

The method of remotely monitoring the health status of data center nodes through IPMI is used to monitor the running status of servers, including CPU, hard disk, fan, memory, power supply, motherboard, etc., and the status of all servers can be remotely monitored through the monitoring terminal, which can be real-time Monitor server load, CPU and memory usage, network fluctuations, etc., and make judgments based on the safety parameter warning line set by the administrator. Send notifications to managers. This method can save a lot of manpower during daily maintenance of server operation, and avoid downtime or other abnormalities in the data center due to various operational hidden dangers. It will greatly reduce the risk of data center node downtime and ensure the reliability of the data center.

Description of drawings

In order to illustrate the technical solution of the present invention more clearly, the accompanying drawings that need to be used in the description will be briefly introduced below. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. As far as people are concerned, other drawings can also be obtained based on these drawings on the premise of not paying creative work.

Fig. 1 is a flow chart of a method for remotely monitoring the health status of data center nodes through IPMI;

Fig. 2 is a schematic diagram of an embodiment of the present invention.

Detailed ways

In order to make the purpose, features and advantages of the present invention more obvious and understandable, the technical solutions protected by the present invention will be clearly and completely described below using specific embodiments and accompanying drawings. Obviously, the implementation described below Examples are only some embodiments of the present invention, but not all embodiments. Based on the embodiments in this patent, all other embodiments obtained by persons of ordinary skill in the art without creative efforts fall within the protection scope of this patent.

This embodiment provides a method for remotely monitoring the health status of node 3 of data center 2 through IPMI, as shown in Figure 1 and Figure 2, monitoring terminal 1, the method includes:

S1, the monitoring terminal 1 and the data center 2 are remotely connected to the same local area network, so that the monitoring terminal 1 communicates with each node 3 of the data center 2;

S2, the monitoring terminal 1 accesses each node 3 of the data center 2 through the IPMI protocol;

S3, the monitoring terminal 1 acquires temperature information, fan information, voltage information, network card working conditions and operating system working conditions of each node 3;

S4, the monitoring terminal 1 compares the obtained information with the corresponding threshold value, and when any of the obtained data information exceeds the threshold value, an alarm message is sent in the data center 2 to prompt maintenance personnel.

In this embodiment, the method includes: the monitoring terminal 1 monitors the temperature information, fan information, voltage information, network card working conditions and operating system working conditions of each node 3 every time a preset time interval passes.

In this embodiment, the monitoring terminal 1 analyzes the received data information, classifies the analyzed data information through corresponding alarm parameters, forms corresponding data information for data exceeding the threshold, and sends alarm information in the data center 2 , to prompt the maintenance personnel, and at the same time, the alarm information is formed into an alarm log;

The monitoring terminal 1 presents the obtained data information and alarm information to the maintenance personnel, and at the same time receives the control instructions from the maintenance personnel to the monitoring terminal 1;

The monitoring terminal 1 provides a unified CUI interface for maintenance personnel to query and browse historical logs and alarm information, and to set alarm parameters.

In this embodiment, the temperature information includes CPU temperature, motherboard temperature, backplane temperature, and hot-swap module temperature;

Voltage information includes CPU voltage, motherboard voltage, SCSI backplane voltage, and hot-swap module voltage

Fan information includes CPU fans, motherboard fans, backplane fans, and hot-swap module fans.

In this embodiment, the monitoring terminal 1 has an I2C device interface, and each node 3 is provided with an I2C Slave interface; the monitoring terminal 1 generates a clock through the I2C device interface to initiate communication to the I2C Slave interface of each node 3, and each node 3 passes through the I2C Slave interface. To respond, the I2C device interface of the monitoring terminal 1 and the I2C Slave interface of each node 3 transmit data through the IPMI protocol.

In the present embodiment, the step monitoring terminal 1 accessing each node 3 of the data center 2 through the IPMI protocol also includes:

The monitoring terminal 1 configures the alarm information, including: node 3 name, node 3 device name, node 3 alarm event name, node 3 alarm description, node 3 alarm trigger value;

The monitoring terminal 1 configures the types of alarm information, including: alarm power supply and distribution, alarm environment, and alarm security;

Monitoring terminal 1 is configured with alarm masking control, node 3 alarm masking, node 3 equipment alarm masking and node 3 alarm event masking, node 3 alarming is masked by configuring the masking mode and masking time period, and configuration release masking control is configured;

The monitoring terminal 1 configures the alarm level, and configures the node 3 level;

Configure node 3 CPU, hard disk, and motherboard alarms as high-level alarms, configure node 3 temperature alarms as intermediate alarms, configure node 3 software alarms as low-level alarms; the specific grading method can be classified according to actual use, and the specific grading method is not limited.

Configure node 3’s power supply and distribution category as high-level alarms, configure node 3’s environment category as medium-level alarms, and configure node 3’s security category as low-level alarms; the specific grading method can be classified according to actual use, and the specific grading method is not limited.

According to the alarm rules, when multiple alarms are generated, the monitoring terminal 1 will give priority to alarms with higher alarm levels according to the configured alarm levels;

Or get multiple alarms in the same time period, according to the configured alarm level, give priority to the alarm with higher alarm level;

When multiple alarms are generated, the low-level alarms are cached for a preset time and sent out;

When a certain node 3 alarms multiple times within the preset time period, it is considered as a high-frequency alarm, and a high-frequency alarm prompt is issued after the preset time period.

In this embodiment, the monitoring terminal 1 screens the alarm keywords, combines the same alarm keywords belonging to the same time interval, and counts the frequency of alarm occurrences, counts the frequency of alarm occurrences according to the time interval, and displays them to maintenance personnel.

In the present embodiment, the step monitoring terminal 1 accessing each node 3 of the data center 2 through the IPMI protocol also includes:

The monitoring terminal 1 encapsulates the configured alarm information, alarm information type, and alarm shielding control information, and sends the encapsulated information to each node 3 of the data center 2

Encapsulation information includes: configure alarm information MAC address layer, configure alarm information data processing terminal IP layer, configure alarm information data frame segment;

In the configuration alarm information data frame segment, when the lowest bit of the frame segment is 0, it indicates the address code of the request/reply response component; when the lowest bit is 1, it indicates the software ID of the request/reply response; the upper 7 bits of the byte Indicates the specific address code and software ID;

The parity information code is set in the configuration alarm information data frame segment. When the parity information code is an even number, it means that the message is a request. When the parity information code is an odd number, it means that the message is a response. When the message is a basic control request Or the byte is 00h and 01h in status response;

In the configuration alarm information data frame segment, there is also a serial number generated by the requester itself, which is used to identify different requests when the monitoring terminal 1 needs to send multiple requests; the serial number is a subcomponent of the component that responds/receives the message number or sub-address number.

The above description of the disclosed embodiments is provided to enable any person 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 present invention will not 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 (8)

1. a kind of method by IPMI remote monitoring data Centroid health status, including：Monitor terminal, its feature exist In method includes：

Monitor terminal is remotely accessed in same local area network with data center, makes monitor terminal node progress each with data center Communication connection；

Monitor terminal accesses each node of data center by IPMI protocol；

Monitor terminal obtains temperature information, fan information, information of voltage, network interface card working condition and the operating system of each node Working condition；

Monitor terminal by the information of acquisition compared with corresponding threshold value, when in the data message of acquisition, having beyond threshold value, Alert in the data center, prompting maintenance personnel.

2. the method according to claim 1 by IPMI remote monitoring data Centroid health status, its feature exists In,

Method includes：

Monitor terminal often passes through a prefixed time interval, to the temperature information of each node, fan information, information of voltage, network interface card Working condition and operating system working condition are monitored.

3. the method according to claim 1 by IPMI remote monitoring data Centroid health status, its feature exists In,

Monitor terminal parses the data message of reception, by and parse after data message by corresponding alarm parameter into Data more than threshold value are formed corresponding data message, in the data center alert, prompting maintenance people by row classification Member, while warning message is formed into alarm log；

The data message got and warning message are presented to maintenance personnel by monitor terminal, while receive maintenance personnel to prison The control instruction of control terminal；

Monitor terminal provides unified CUI interfaces, for inquiry of the maintenance personnel to history log, warning message and browse and Alarm parameters are configured.

4. the method according to claim 1 by IPMI remote monitoring data Centroid health status, its feature exists In,

Temperature information includes cpu temperature, mainboard temperature, backboard temperature, hot plug module temperature；

Information of voltage includes CPU voltages, mainboard voltage, SCSI backboard voltages, hot plug module voltage

Fan information includes cpu fan, mainboard fan, backboard fan, hot plug module fan.

5. the method according to claim 1 by IPMI remote monitoring data Centroid health status, its feature exists In,

Monitor terminal has I2C equipment interfaces, and each node is provided with I2C Slave interfaces；

The I2C Slave interfaces of each node of clockwise initiate communication, each node when monitor terminal is produced by I2C equipment interfaces Responded by I2C Slave interfaces

The I2C equipment interfaces of monitor terminal transmit data with the I2C Slave interfaces of each node by IPMI protocol.

6. the method according to claim 1 by IPMI remote monitoring data Centroid health status, its feature exists In,

Each node that step monitor terminal accesses data center by IPMI protocol further includes：

Monitor terminal configures warning message, including：Nodename, node device title, node alert event title, section Point alarm description, node alarm trigger value；

Monitor terminal configures warning message type, including：Power supply and distribution of alarming class, alarm environmental classes, alarm security type；

Monitor terminal configuration alarm shielding control, node alarm shielding, node device alarm shielding and the shielding of node alert event, Alarmed by configuring shielding mode and shielding period come masked nodes, configuration releases shielding control；

Monitor terminal configures alert levels, configuration node rank；

Configuration node CPU, hard disk, the alarm of mainboard is advanced alarm, and configuration node temperature alarming is alarmed for middle rank, configuration node Software class alarm is rudimentary alarm；

Configuration node power supply and distribution class is advanced alarm, and the alarm of configuration node environmental classes is middle rank alarm, and configuration node security type is Rudimentary alarm；

Monitor terminal, when multiple alarms produce, according to the alert levels of configuration, preferentially sends alert levels according to alarm rule High alarm；

Or multiple alarms are obtained in the same period, according to the alert levels of configuration, preferentially send the high alarm of alert levels；

When multiple alarms produce, after the low alarm caching preset time of rank, send；

When certain node is repeatedly alarmed in preset time period, then it is assumed that be high frequency time alarm, sent after preset time period High frequency time alarm.

7. the method according to claim 1 by IPMI remote monitoring data Centroid health status, its feature exists In,

Monitor terminal screening alarm keyword, will belong in same time interval, and Same Alarm keyword merges, statistics alarm hair Raw frequency, the frequency of alarm generation is counted according to time interval, shows maintenance personnel.

8. the method according to claim 6 by IPMI remote monitoring data Centroid health status, its feature exists In,

Each node that step monitor terminal accesses data center by IPMI protocol further includes：

Monitor terminal will be packaged with the warning message postponed, warning message type, alarm shielding control information, form encapsulation Information is sent to each node of data center

Packaging information includes：Warning message MAC Address layer is configured, configures IP layers of warning message data processing terminal, configuration alarm Information data frame section；

Configure the address for the component for representing request/reply response in warning message data frame section when the lowest order of the frame section is 0 Code, the software I D of request/reply response is represented when lowest order is 1；High 7 of the byte represent specific address code and software ID；

Parity information is equipped with warning message data frame section is configured to encode, this message is represented when parity information is encoded to even number It is a request, it is a response that this message is represented when parity information is encoded to odd number；When message is a basic control The byte is 00h and 01h when request or condition responsive；

The sequence number that requestor voluntarily generates is additionally provided with warning message data frame section is configured, when monitor terminal needs to send Identifying different requests during a plurality of request；Sequence number is in response to/receives the subgroup piece number or subaddressing of the component of the message Number.