CN107682222A - A kind of information monitoring method, apparatus and system - Google Patents

Abstract

The present invention provides a kind of information monitoring method and device, the above method and comprised the following steps：Client obtains the characteristic information of the monitored item of monitored node by management node；The client is according to the characteristic information of the monitored item, it is determined that corresponding display mode.In above-mentioned technical proposal, the characteristic information of the automatic monitored item for obtaining and showing monitored node is realized, simplifies operating process, and the monitored results obtained can intuitively be shown so that monitoring efficiency greatly promotes, and improves Consumer's Experience.

Description

An information monitoring method, device and system

technical field

The invention belongs to the technical field of computers, and in particular relates to an information monitoring method, device and system.

Background technique

A cluster is a computer cluster system with high performance, high scalability, and high cost performance with a single system image that is formed by interconnecting a group of computer systems through a high-performance network or local network. With the widespread application of the cluster system in scientific computing, business operations, etc., the role of the cluster system is becoming more and more important, and it has gradually become an indispensable tool in the above fields. However, the number of nodes in the cluster system is huge. Hardware information, such as CPU utilization, memory, load, etc., requires complex and repetitive operations (for example, you need to manually add monitoring items and path selection when adding), and the obtained monitoring results cannot be displayed intuitively, which is both It brings inconvenience to the user and greatly reduces the monitoring efficiency, seriously affecting the user experience.

Therefore, it is urgent to provide an information monitoring solution to solve the above technical problems.

Contents of the invention

The present invention provides an information monitoring method, device and system to solve the above problems.

An embodiment of the present invention provides an information monitoring method, including the following steps: the client obtains the feature information of the monitored item of the monitored node through the management node;

The client determines a corresponding display mode according to the feature information of the monitoring item.

An embodiment of the present invention also provides an information monitoring device, including a processor, adapted to implement various instructions; a storage device, adapted to store multiple instructions, and the instructions are adapted to be loaded and executed by the processor;

The client obtains the characteristic information of the monitored items of the monitored node through the management node;

The client determines a corresponding display mode according to the feature information of the monitoring item.

An embodiment of the present invention also provides an information monitoring system, including a client, a management node, and one or more monitored nodes, wherein the client is connected to the one or more monitored nodes through the management node.

The technical solution provided by the embodiment of the present invention: the client obtains the characteristic information of the monitored item of the monitored node through the management node; the client determines the corresponding display mode according to the characteristic information of the monitored item.

In the above technical solution, the feature information of the monitoring item of the monitored node is automatically obtained and displayed, the operation process is simplified, and the obtained monitoring results can be displayed intuitively, so that the monitoring efficiency is greatly improved, and the user experience is improved.

Description of drawings

The accompanying drawings described here are used to provide a further understanding of the present invention and constitute a part of the application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention. In the attached picture:

Fig. 1 shows the flow chart of the information monitoring method of embodiment 1 of the present invention;

FIG. 2 is a structural diagram of an information monitoring device according to Embodiment 2 of the present invention;

Fig. 3 shows the structural diagram of the information monitoring system of Embodiment 3 of the present invention;

FIG. 4 is a structural diagram of an information monitoring system according to Embodiment 4 of the present invention.

detailed description

Hereinafter, the present invention will be described in detail with reference to the drawings and examples. It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

The invention integrates the web plug-in and the underlying monitoring program, stores the monitoring data in the database, and captures the monitoring data from the database in real time or at regular intervals, thereby achieving dynamic monitoring of the GPU cluster.

Working principle: The client adopts the B/S architecture management system, monitors the overall situation of the cluster GPU card through the web page, can select a certain GPU card of a monitored node and some monitoring items, and send HTTP requests to the management node; management After the node accepts the HTTP request, it accesses the monitoring data in the database and returns it to the web browser; the browser receives the monitoring data sent by the management node, displays the monitoring data in the form of icons, and dynamically refreshes the page to dynamically monitor the status of each GPU card in the cluster. feature data.

Among them, the management node and each computing node share the monitoring configuration file; the computing node regularly collects monitoring data according to the monitoring configuration file and sends it to the management node;

The management node receives the monitoring data, sorts and saves it to the database;

When the front-end page sends a data request, the monitoring data in the database is returned to the browser;

The browser displays the returned results.

The above solution has the following advantages: it can monitor the overall running status of the cluster, dynamically configure the monitoring items to be monitored through the monitoring configuration file, and can complete the unified monitoring of each GPU node in a large-scale cluster, which is convenient and practical for cluster use wide.

in particular:

When the page is opened for monitoring and viewing, the front-end page will regularly send Http requests to the server to view the overall information. When viewing certain monitoring items of a monitored node, it will send json data similar to the following to the management node {node: "node1", gpu: "gpu1", config: "cpu, mem, load"};

The management node receives the browser request, processes the request, and sends commands to the database; receives the monitoring data returned by the database, encapsulates the monitoring data and returns it to the browser for display.

Wherein, the management node starts the server-side monitoring service. If it is started for the first time, a new database table is created through the monitoring configuration file, and the socker communication is opened to wait for the monitoring data reported by the monitored node; it is also used for processing after receiving the monitoring data.

The monitored node starts the client service and reads the connection information sent by the management node. After a successful connection, it will obtain the monitoring data of the monitoring item (eg GPU card) through the nvidia programming interface, assemble and package the monitoring data and send it to the management node.

After the management node receives the monitoring data, it packs the monitoring data into the database as a whole. When the page sends a request, it returns the monitoring data to the browser, and the browser displays the monitoring data in real time through the icon plug-in.

Fig. 1 shows the flow chart of the information monitoring method of embodiment 1 of the present invention, comprises the following steps:

Step 101: the client obtains the feature information of the monitoring item of the monitored node through the management node;

Further, the monitored items include: GPU, hard disk, and memory; the characteristic information includes: utilization rate, abnormal information, and alarm information.

Further, before the client obtains the feature information of the monitored item of the monitored node through the management node, it also includes:

The management node acquires feature information of monitored items from the monitored nodes and stores the feature information in a database.

Preferably, the database is located in the management node; or, the database is deployed independently of the management node.

Further, the client sends a monitoring data acquisition request to the management node, wherein the monitoring data acquisition request carries monitoring item information;

After receiving the monitor data acquisition request, the management node acquires feature information corresponding to the monitor item from a database and transmits the feature information to the client.

Further, the management node and the monitored node store monitoring configuration files, wherein the monitoring configuration files include monitoring items and monitoring periods.

Preferably, the monitored node reports the characteristic information of the monitored item to the management node according to the monitoring period.

Step 102: The client determines a corresponding display mode according to the characteristic information of the monitoring item.

Further, if there are multiple monitoring items, display the characteristic information of multiple monitoring items in the first display mode; or,

If there is one monitoring item, display the feature information of the monitoring item in a second display manner.

Wherein, the first display mode includes the name of each monitoring item, the utilization rate of each monitoring item, and the storage space occupancy rate of each monitoring item;

The second display mode includes: the name of the monitored node where the monitoring item is located, the address of the monitored node where the monitoring item is located, the system architecture, the operating system, the utilization rate of the monitoring item in different periods, and the storage space occupancy rate of the monitoring item in different periods.

Of course, the information displayed above is only for illustration, and other information may also be added and displayed, which is not used to limit the scope of protection of this embodiment.

The above solution has a clear structure, is easy to use, easy to maintain, and convenient to monitor each monitored node in the GPU cluster, freeing the cluster administrator from complicated work, and conveniently realizes the display of multiple GPU nodes.

Fig. 2 is a structural diagram of an information monitoring device according to Embodiment 2 of the present invention, including a processor adapted to implement instructions; a storage device adapted to store a plurality of instructions adapted to be loaded and executed by the processor ;

The client obtains the characteristic information of the monitored items of the monitored node through the management node;

The client determines a corresponding display mode according to the feature information of the monitoring item.

Further, the monitored items include: GPU, hard disk, and memory; the status feature information includes: utilization rate, abnormal information, and alarm information.

Further, before the client obtains the feature information of the monitored item of the monitored node through the management node, it also includes:

The management node acquires feature information of monitored items from the monitored nodes and stores the feature information in a database.

Preferably, the database is located in the management node; or, the database is deployed independently of the management node.

Further, the client sends a monitoring data acquisition request to the management node, wherein the monitoring data acquisition request carries monitoring item information;

After receiving the monitor data acquisition request, the management node acquires feature information corresponding to the monitor item from a database and transmits the feature information to the client.

Further, the management node and the monitored node store monitoring configuration files, wherein the monitoring configuration files include monitoring items and monitoring periods.

Preferably, the monitored node reports the characteristic information of the monitored item to the management node according to the monitoring period.

The software installation directory of the management node and the monitored node should be shared. The management node can dynamically obtain the configuration items and the information of the monitored node through the monitoring configuration file. The management node and the monitored node should be able to communicate with each other.

Further, if there are multiple monitoring items, display the characteristic information of multiple monitoring items in the first display mode; or,

If there is one monitoring item, display the feature information of the monitoring item in a second display manner.

Wherein, the first display mode includes the name of each monitoring item, the utilization rate of each monitoring item, and the storage space occupancy rate of each monitoring item;

The second display mode includes: the name of the monitored node where the monitoring item is located, the address of the monitored node where the monitoring item is located, the system architecture, the operating system, the utilization rate of the monitoring item in different periods, and the storage space occupancy rate of the monitoring item in different periods.

Figure 3 is a structural diagram of an information monitoring system according to Embodiment 3 of the present invention, including a client, a management node, and one or more monitored nodes, wherein the client communicates with the one or more monitored nodes through the management node The monitored node is connected.

Further, the client is used to obtain feature information of monitored items of the monitored node through the management node; and is also used to determine a corresponding display mode according to the feature information of the monitored items.

Further, the management node is used to obtain the characteristic information of the monitored item from the monitored node and store the characteristic information in the database; Monitor the characteristic information corresponding to the item and transmit the characteristic information to the client; and also receive the characteristic information of the monitored item reported by the monitored node.

Further, the monitored node is configured to report the characteristic information of the monitored item to the management node according to the monitoring cycle.

In the above solution, the administrator accesses the web interface through a browser, and intuitively monitors the operation status of each monitored node and each GPU card in the entire cluster, and can display the real-time monitoring information of resources and alarms, which is helpful for administrators and users. It is of great help to daily work, provides guarantee for the stable operation of GPU nodes, and has strong applicability to GPU cluster users and GPU cluster managers.

Fig. 4 shows the structure diagram of the information monitoring system of embodiment 4 of the present invention, including client; Management node; Monitored node 1 and corresponding monitoring item GPU1; Monitored node 2 and corresponding monitoring item GPU2; Monitored node n And the corresponding monitoring item GPUn.

The client is used to obtain the feature information of the GPU1 of the monitored node 1, the feature information of the GPU2 of the monitored node 2, and the feature information of the GPUn of the monitored node n through the management node; Feature information to determine the corresponding display mode.

Further, the management node is used to obtain the feature information of the GPU1 of the monitored node 1, the feature information of the GPU2 of the monitored node 2, and the feature information of the GPUn of the monitored node n from the monitored nodes, and the The feature information is stored in the database;

It is also used to obtain the characteristic information corresponding to the monitored item from the database after the monitoring data acquisition request sent by the client and transmit the characteristic information to the client; it is also used to receive the monitoring data reported by the monitored node The feature information of the item.

Further, the monitored node is configured to report the characteristic information of the monitored item to the management node according to the monitoring cycle.

The technical solution provided by the embodiment of the present invention: the client obtains the characteristic information of the monitored item of the monitored node through the management node; the client determines the corresponding display mode according to the characteristic information of the monitored item.

In the above technical solution, the feature information of the monitoring item of the monitored node is automatically obtained and displayed, the operation process is simplified, and the obtained monitoring results can be displayed intuitively, so that the monitoring efficiency is greatly improved, and the user experience is improved.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (21)

1. A kind of information monitoring method, is characterized in that, comprises the following steps: The client obtains the characteristic information of the monitored items of the monitored node through the management node; The client determines a corresponding display mode according to the feature information of the monitoring item. 2 . The information monitoring method according to claim 1 , wherein the monitoring items include: GPU, hard disk, and memory; and the feature information includes: utilization rate, abnormal information, and alarm information. 3. The information monitoring method according to claim 1 or 2, wherein the client further includes: The management node acquires feature information of monitored items from the monitored nodes and stores the feature information in a database. 4. The information monitoring method according to claim 3, wherein the database is located in the management node; or, the database is deployed independently of the management node. 5. The information monitoring method according to claim 4, wherein the client sends a monitoring data acquisition request to the management node, wherein the monitoring data acquisition request carries monitoring item information; After receiving the monitor data acquisition request, the management node acquires feature information corresponding to the monitor item from a database and transmits the feature information to the client. 6. The information monitoring method according to claim 5, wherein the management node and the monitored node store monitoring configuration files, wherein the monitoring configuration files include monitoring items and monitoring periods. 7. The information monitoring method according to claim 6, wherein the monitored node reports the characteristic information of the monitored item to the management node according to the monitoring cycle. 8. The information monitoring method according to claim 1, wherein if there are multiple monitoring items, the feature information of the multiple monitoring items is displayed in a first display mode; or, If there is one monitoring item, display the feature information of the monitoring item in a second display manner. 9. The information monitoring method according to claim 8, wherein the first display mode includes the name of each monitoring item, the utilization rate of each monitoring item, and the storage space occupancy rate of each monitoring item; The second display mode includes: the name of the monitored node where the monitoring item is located, the address of the monitored node where the monitoring item is located, the system architecture, the operating system, the utilization rate of the monitoring item in different periods, and the storage space occupancy rate of the monitoring item in different periods. 10. An information monitoring device, characterized in that it comprises a processor adapted to implement instructions; a storage device adapted to store a plurality of instructions, and the instructions are adapted to be loaded and executed by the processor; The client obtains the characteristic information of the monitored items of the monitored node through the management node; The client determines a corresponding display mode according to the feature information of the monitoring item. 11 . The information monitoring device according to claim 10 , wherein the monitoring items include: GPU, hard disk, and memory; and the feature information includes: utilization rate, abnormal information, and alarm information. 12. The information monitoring device according to claim 10 or 11, wherein, before the client acquires the feature information of the monitored item of the monitored node through the management node, it also includes: The management node acquires feature information of monitored items from the monitored nodes and stores the feature information in a database. 13. The information monitoring device according to claim 12, wherein the database is located in the management node; or, the database is deployed independently of the management node. 14. The information monitoring device according to claim 13, wherein the client sends a monitoring data acquisition request to the management node, wherein the monitoring data acquisition request carries monitoring item information; After receiving the monitor data acquisition request, the management node acquires feature information corresponding to the monitor item from a database and transmits the feature information to the client. 15. The information monitoring device according to claim 14, wherein the management node and the monitored node store monitoring configuration files, wherein the monitoring configuration files include monitoring items and monitoring periods. 16. The information monitoring device according to claim 15, wherein the monitored node reports the characteristic information of the monitored item to the management node according to the monitoring cycle. 17. The information monitoring device according to claim 10, wherein if there are multiple monitoring items, the feature information of multiple monitoring items is displayed in a first display manner; or, If there is one monitoring item, display the feature information of the monitoring item in a second display manner. 18. An information monitoring system, characterized by comprising a client, a management node, and one or more monitored nodes, wherein the client is connected to the one or more monitored nodes through the management node. 19. The information monitoring system according to claim 18, wherein the client is configured to acquire the feature information of the monitored item of the monitored node through the management node; it is also used to, according to the feature information of the monitored item, Determine the corresponding display mode. 20. The information monitoring system according to claim 18, characterized in that, the management node is used to acquire the characteristic information of the monitored item from the monitored node and store the characteristic information in the database; After obtaining the monitoring data acquisition request sent by the terminal, the feature information corresponding to the monitoring item is obtained from the database and the feature information is transmitted to the client; it is also used to receive the feature information of the monitoring item reported by the monitored node. 21. The information monitoring system according to claim 18, wherein the monitored node is configured to report the characteristic information of the monitored item to the management node according to a monitoring cycle.