CN1266918C

CN1266918C - Method and device for machine group system monitoring

Info

Publication number: CN1266918C
Application number: CN 03146287
Authority: CN
Inventors: 程菊生; 崔吉顺; 彭涛; 田宏萍; 金正操; 蒋晓华
Original assignee: Lenovo Beijing Ltd
Current assignee: Lenovo Beijing Ltd
Priority date: 2003-07-08
Filing date: 2003-07-08
Publication date: 2006-07-26
Anticipated expiration: 2023-07-08
Also published as: CN1476224A

Abstract

The present invention provides a device for monitoring cluster systems. The present invention comprises a server, a wireless network bridge connecting device, a user wireless receiving and transmitting device and user devices, wherein the server is used for storing the data of a cluster monitoring system and managing the cluster monitoring system; the wireless network bridge connecting device is coupled with the server so as to get access to the cluster monitoring system in a wireless mode; the user wireless receiving and transmitting device is coupled with the wireless network bridge connecting device through a wireless link; the user devices are respectively coupled with the server through the user wireless receiving and transmitting device. Besides, the server comprises a processing device; according to the requirements of the user device, the data of the cluster monitoring system is converted to information according to a communication protocol used by a local area network device, and the information meets the data format of the wireless network bridge connecting device and the user wireless receiving and transmitting device. By the present invention, users can know the operational conditions of clusters at any place and at any time and can monitor and manage the cluster systems at any place and at any time.

Description

Method and device for monitoring cluster system

Technical Field

The invention relates to computer group monitoring, in particular to a method and a device for monitoring and alarming a computer group system; more particularly, it relates to a computer cluster monitoring by PDA and cluster alarming by short message function of mobile phone.

Background

A cluster system is an aggregation of multiple server (node) machines. The servers are aggregated together over a private high-speed network to form a single-image super server through a software system such as an middleware. In the running process of the cluster system, the health state and the running state of hardware of the cluster system need to be monitored at any time. Therefore, a user (administrator) can timely know the running state of the cluster, timely find abnormal conditions occurring in the running process, and perform corresponding processing to ensure the normal running of the cluster system. Each node has its own memory, I/O devices and operating system. The node machines are connected together through a high-speed network, and form a super server under the cooperation of middleware and the like. The cluster server plays an important role in large-scale scientific calculation, such as petroleum geology and the like.

In a cluster monitoring system, it is often necessary to monitor the operating status of each node of the cluster. The conventional cluster monitoring and management system adopts a console mode, and when abnormal conditions occur to the nodes, such as temperature exceeding or falling below a set alarm threshold, the cluster generates an alarm event. Generally speaking, these alarm events are displayed through the monitoring interface of the cluster monitoring system or prompted by different sounds, so that a cluster user can know the alarm condition of a certain node in the cluster as long as he or she views the alarm interface of the monitoring system or hears the sounds, and for the node with serious alarm, he or she can perform corresponding node maintenance according to the alarm event. For a cluster user, if a cluster is maintained, the alarm interface of the monitoring system needs to be continuously checked so as to find a serious alarm event in time and further maintain the cluster in time, which is undoubtedly a huge constraint for the monitoring user. Therefore, a convenient monitoring method is needed, which can timely acquire the monitoring information of the cluster, especially the serious alarm information in the alarm condition, when the cluster user leaves the cluster monitoring environment, so as to take corresponding measures to solve the problem.

The invention realizes remote monitoring of the cluster by utilizing the mobile communication technology, and the monitoring alarm information of the cluster is sent to the mobile phone of a cluster user by short messages, so that the user can know the running condition of the cluster at any time and any place, the remote wireless monitoring of the cluster is realized by sending the monitoring information to the PDA, the monitoring management of the cluster system is realized by accessing a web address through the PDA, and the user can realize the monitoring management of the cluster system at any time and any place.

Disclosure of Invention

The invention aims to provide a method and a device for monitoring a cluster system, aiming at the monitoring problem in the prior art of the cluster system, the monitoring alarm information of the cluster is sent to a mobile phone of a cluster user by a short message, the user can know the running condition of the cluster at any time and any place, the monitoring information is sent to a PDA to realize remote wireless monitoring of the cluster, the monitoring management of the cluster system is realized by accessing a web address through the PDA, and the user can realize the monitoring management of the cluster system at any time and any place.

The invention provides a device for monitoring a cluster system, which comprises:

the server is positioned in the cluster system and used for collecting and storing the monitoring data of each node in the cluster monitoring system; converting the stored monitoring data into a data format meeting the requirements of the wireless network bridge connecting device and the user wireless transceiver device according to the requirements of the user device and a communication protocol transmitted to the wireless network bridge connecting device, and transmitting the data format to the wireless network bridge connecting device;

the wireless network bridge connecting device is connected with the server and used for sending the monitoring data transmitted by the server to the user wireless transceiver in a wireless mode;

the user wireless transceiver is coupled with the wireless network bridge connection device through a wireless link and used for receiving the monitoring data sent by the wireless network bridge connection device and transmitting the monitoring data to the user device;

and the user device is connected with the user wireless transceiver and used for receiving the monitoring data of the user wireless transceiver and displaying the monitoring data to the user.

Optionally, the device further includes a local area network device, respectively connected to the server and the wireless network bridge connection device, and configured to implement connection between the server and the wireless network bridge connection device.

Preferably, the server includes: a basic service provider BSP device module and a database;

a Basic Service Provider (BSP) device module, configured to implement monitoring data of each node in the cluster system;

the database is used for realizing the monitoring data of the node machine acquired by the storage BSP module;

the device also comprises a configuration device and an interface device; wherein,

the configuration device is used for configuring the information customized by the cluster user and caching the customized information;

and the interface device is used for screening the monitoring data of the node machine acquired by the BSP module according to the customized information cached by the configuration device and then writing the monitoring data into the database.

Optionally, the configuration device includes a short message reminding device, which is used for realizing an interface between the BSP and the wireless network bridge connection device, and filtering short message information customized by a group user; the wireless network connection device comprises a controllable device, and is used for sending a data packet of the short message information through a Socket (Socket) so as to be processed into a wirelessly sent short message by a pubcfd (public configuration) program in the controllable device.

Preferably, the local area network device includes: local area network: coupled to the server, the data transmission is realized based on a TCP/IP protocol; HUB (HUB): a connection for managing a local area network; a network controller: and the management module is used for managing remote login management of the server through the local area network and the HUB.

Optionally, the user wireless transceiver includes a PDA wireless network card, which is used for communicating the user's PDA with the wireless network bridge device; the user device is a PDA, and the PDA wireless network card is integrated in the PDA.

Preferably, the server includes a data dividing device, configured to classify the collected monitoring data of each node according to the rack or the cluster; and the display processing device is used for processing the classified monitoring data into a data format which is convenient to display in a histogram.

The invention also provides a method for monitoring the cluster system, which comprises the following steps:

presetting a preset condition for a user to acquire monitoring data;

a server in the cluster system acquires monitoring data of each node of the cluster system; and storing the acquired monitoring data;

and screening the acquired or stored monitoring data of the cluster system according to the set preset condition for acquiring the monitoring data by the user, and transmitting the screened monitoring data to the user in a wireless manner.

Preferably, the step of transmitting the filtered monitoring data to the user includes: the screened monitoring data is transmitted to the user in a manner that the information is displayed in a bar graph supported by the PDA.

Optionally, the step of transmitting the stored monitoring data to a user comprises: after the JSP receives a message of requesting to send information from a user, the JSP generates query content according to the set data format and the preset conditions, and transmits the query content to SERVLET/JSP so as to call an interface provided by JAVABEAN to query a database; then processing and recombining the inquired data by SERVLET/JSP to enable the inquired data to generate a dynamic webpage; the JSP organizes the HTML language display interface for display by the PDA.

Optionally, the step of transmitting the filtered monitoring data to the user includes: transmitting the screened monitoring data to a user in a short message format; the predetermined conditions include fleet catastrophic alarm conditions and/or general alarm conditions and/or minor fault alarm conditions.

Preferably, the step of acquiring the monitoring data of the cluster system includes: the BSP provided by the server acquires monitoring data of each node machine through a local area network and a serial port;

the step of storing the acquired monitoring data comprises: writing the acquired monitoring data into a database through an API 0 (application program interface 0) to maintain the monitoring data;

said step of transmitting said converted monitoring data to a user and said step of converting said stored monitoring data according to said set predetermined condition for the user to obtain monitoring data comprises comparing said monitoring data to a predetermined threshold in the predetermined condition and generating an alarm event if said predetermined threshold is exceeded; the alarm event is sent through API 2 (application program interface 2).

The invention can make user know the running condition of cluster at any time and any place, and realize the monitoring and management of cluster system at any time and any place.

Drawings

FIG. 1 illustrates a functional block diagram of a processing device in a monitoring device utilizing a PDA (personal digital Assistant) in a fleet monitoring system in accordance with an embodiment of the present invention;

FIG. 2 shows a functional block diagram of a monitoring device using a PDA according to an embodiment of the present invention;

FIG. 3 depicts a functional block diagram of a cluster system monitored using a cellular telephone according to another embodiment of the present invention;

fig. 4 depicts a schematic diagram of a short message sending device using a cellular phone short message monitoring cluster system according to an embodiment of the present invention, and depicts a flow of an implementation method through these devices.

Detailed Description

To facilitate understanding and implementing the invention by those of ordinary skill in the art, embodiments of the invention are now described with reference to the drawings.

Fig. 1 shows a functional block diagram of a processing device in a monitoring device utilizing a PDA (personal digital assistant) in a fleet monitoring system in accordance with an embodiment of the present invention.

In the embodiment of the present invention, the monitoring interface of the PDA is implemented by using a WEB browsing interface, and using Java Servlet (Java Servlet is a service program provided by Java to provide request and response services on the WEB) and JSP technology (JSP, JavaServer Page, server-side WEB Page program, which controls the content of a WEB Page by using a program working on the server side), obtains data, performs analysis processing, and displays the data to the client browser through a graphical interface. At the server end, PDA logs in the server through wireless link, which is basically consistent with the mode of computer logging in, PDA logs in the server of the cluster monitoring system through wireless Internet or the mode of direct server logging in. Content on Java Servlet and JSP is referable

www.socent.com/programming/category.aspcate＝14。

A user interacts with the SERVLET/JSP through the JSP by sending a login request, accesses a database through a JAVA BEAN (the JavaBean is a Java-based software component program, and realizes a Web-based file uploading function in the JSP), and the JSP also directly interfaces with the JAVA BEAN.

The Servlet processes and dispatches the request, processes and reorganizes the data information, and enables the data information to generate the dynamic webpage.

And the JAVA Beans acquire all cluster data information from the database and provide an interface for inquiring the database.

SERVLET/JSP and JSP, convert data into statistical graphs (histograms, etc.), organize HTML language display interfaces by tables, etc., and store information in clients.

The data base stores various related information and data of the cluster, and the data are processed by the device and then monitored by the PDA.

The processing flow is that the cluster monitoring system stores the collected monitoring data in a database; generating query contents according to different page requests according to a user query request JSP, transmitting the information to SERVLET/JSP, and calling an interface query database provided by JAVABEAN by the SERVLET/JSP; then filtering the information, and returning the information to the SERVLET/JSP, wherein the SERVLET/JSP processes and recombines the information to generate a dynamic webpage; and finally, the JSP organizes an HTML language display interface and stores the monitoring information in the PDA terminal for display.

Aiming at the characteristics of small display screen and limited display content of the PDA, the monitoring information displayed in the PDA interface is the main information in the monitoring system, and focuses on the cluster maintenance related information concerned by the cluster user.

In an embodiment of the present invention, the following method is used for display and processing in the PDA interface.

Data division: the sub-parameters are displayed on the PDA in a multi-page mode by taking a cabinet (node data) or a cluster (cabinet data) as a unit;

and (4) visually displaying the histogram: in the embodiment of the present invention, due to the display characteristics of the PDA, the monitoring data is processed into a histogram for convenient display; in general, the histogram in a single page of the PDA can clearly display the data of 40 nodes without dragging laterally.

The following good results were obtained on the PDA display:

1) the standard MVC mode (MODEL/VALUE/CONTROL) is flexible and efficient;

2) vivid and visual data display (histogram), convenient real-time monitoring and quick fault location

Fig. 2 shows a schematic block diagram in a monitoring apparatus using a PDA according to an embodiment of the present invention. The monitoring device of the embodiment of the invention comprises the following devices:

a Server: the server can be a dedicated node of the cluster or a server for storing all data of the cluster monitoring system. The server can collect data of each node of the cluster according to a certain time interval, process the collected data and display the data through a connected display. The user can directly log in the server to manage the cluster monitoring system, or can log in the server through the PC connected with the server, or can log in the server through the PDA. The login procedure is described above in connection with fig. 1.

Local area network: the network of the login server adopts a local area network, and a TCP/IP protocol is operated to transmit data.

HUB: realizing the connection of the local area network;

a PC machine: the remote login management can be carried out on the cluster monitoring management server through the network.

Wireless network bridge connector: in this embodiment, an associatively produced Tiansuo LX-AP1500 is used to implement wireless access to a local area network.

PDA wireless network card: in this embodiment, an 802.11B CF wireless network card from Symbol is used, and the maximum flow rate is 11 Mbps.

PDA: the method is used for browsing and monitoring the monitoring system.

The user opens the IE browser on the PDA, inputs the IP address of the monitoring system server, the PDA and the server are connected through the wireless network card, the wireless network bridge connector and the HUB, and data transmission and exchange are carried out through the local area network, so that the user can remotely manage and monitor the cluster.

FIG. 3 depicts a functional block diagram of a fleet system monitored using cellular phone SMS in accordance with another embodiment of the present invention. Mainly include four modules: the system comprises a monitoring host, a wireless conversion board card, a wireless sending module and a user mobile phone. Their function is as follows:

the monitoring host computer: the alarm event generated by the cluster monitoring system is stored in the monitoring host, and the short message customization function is realized in the monitoring host according to the characteristics of the short message;

wireless conversion integrated circuit board: the monitoring host is arranged on a PCI slot of the monitoring host, is used for converting the alarm information in the monitoring host into a format required by the wireless sending module, and is a bridge between the monitoring host and the wireless sending module;

a wireless transmission module: sending the alarm information in a wireless mode, wherein the hardware module adopts a wing GPRS of associative movement;

the mobile phone of the user: the wireless GPRS mobile phone has the function of receiving the short message of the Chinese character system.

Fig. 4 depicts a schematic diagram of a short message sending device using a cellular phone short message monitoring cluster system according to an embodiment of the present invention, and depicts a flow of an implementation method through these devices. The device comprises the following functional modules.

1) A BSP module: the Basic Service Provider realizes data acquisition of each node in the cluster, and the acquired data is the aforementioned 29 types of data. The BSP module collects data from each node of the cluster at regular intervals, typically 30 seconds, for all cluster nodes in a round. Meanwhile, the module can also send the information of the control cluster node of the upper layer interface to each node machine for execution, such as the stop of an operating system, the start and the shutdown of the node machine and other operations, which need to be transmitted and completed by the BSP module.

2) A database: the data of the node machine collected by the BSP module is stored in a database, and the database processes the data and supplies the data to upper interface software for calling.

3) Short message configuration information caching: the short message customization information is cached;

4) a short message reminding module: the interface between the BSP and the short message sending module filters the sent alarm short message according to the content customized by the short message;

5) pubconfd program: the short message reminding module sends a short message data packet to a pubcfd program through Socket;

6) API 0: after the alarm information is generated, the BSP operates the database and writes the alarm information into the database;

7) API 1: the short message reminding module acquires short message customization information from a database;

8) API 2: the application program interface is an interface provided for the BSP to send the short credit and is a calling interface for the short message sending data;

in this embodiment, the work flow of sending the short message is as follows:

1) when the BSP is started, monitoring data are acquired through a local area network and a serial port, the acquired information is written into a database through an API 0 (application program interface 0), and the monitoring information is maintained in real time;

2) meanwhile, comparing the acquired data with a threshold value thereof, and if the acquired data exceeds the range, generating an alarm event;

3) writing the alarm event into a database through an API 0, and simultaneously sending the alarm event to a short message reminding module through an API 2;

4) the short message reminding module acquires short message customized information from a database through an API 1 and stores the short message customized information into a short message customized information cache;

5) the short message customizing module screens and sends short messages through short message customizing information and sends the information to the short message sending module and the wireless module;

6) the short message sending module and the wireless module send out the short message through system hardware.

The following describes these flows in detail, including the flow of sending short messages.

The short message alarm process comprises two steps:

1) short message reservation;

2) and sending the short message after the corresponding event is generated.

1) Short message reservation flow:

the short message reservation refers to that a user can set the receiving level of the short message without receiving the short message through an interface.

After a certain user logs in a monitoring system interface, the menu of 'system configuration- > mobile phone short message configuration' is provided with the functions of not receiving short messages, receiving short messages of which levels and types, and configuring the mobile phone number of the logged-in user.

In an embodiment of the invention, the short message is scheduled into three levels:

a level: and (5) alarming catastrophically. A case where the system is about to stop or has stopped operating, such as a cpu fan stopping rotating, a hard disk remaining space of 0, or the like;

b stage: and (4) generally alarming. The system has serious faults but the condition of continuing operation is not influenced;

c level: light fault alarm

Then, establishing a user and short message level association database table, storing the ID number of the user customizing the information and the preset short message level: the specific procedure is as follows:

create table admin_smsLevel{

adminID int (10) default null, # user ID number, primary key (default null)

Sms _ level varchar (1) default null # predetermined SMS level

}；

2) Process for sending short message after corresponding event is generated

After the corresponding event needing to send the short message is generated, each system sends a request data packet for sending the short message to a pubcfd program, the pubcfd analyzes the content and the type of the short message information to be sent and the node for generating the event from the data packet after receiving the request data packet, analyzes the corresponding users to send the short message according to the node for generating the event, obtains the mobile phone number of the user, checks whether the user reserves the short message or not, sets to receive the short message or not, and finally writes the information into a database and marks that the information needs to be actually sent at last.

The next step is that the program which actually sends the short message regularly inquires the database, and if the short message to be sent is found in the database, the short message is taken out and sent.

The sent short message information is written into the database, the short messages written into the database to be sent are all the short messages to be actually sent, a short message log is formed, and records in the log can be inquired and cleared through inquiring/deleting the short message log on the interface.

The program for describing the database table structure of the short message log comprises the following steps:

create table sms_log{

log _ ID int (10) default 0, # Log record ID, Primary Key

admin _ name varchar (100) default null, # user id

sender char (50) default null, # message origin

sms _ level varchar (1) default null # sms level

message varchar (1) default nul1 # message

got _ time varchar (20) default null # acceptance time

send_type int(1)default null

Wherein: # 1-indicates that it has actually been sent (written to the outbreak pool)

#2- (not actually sent) because the user SMS is not open;

#3- (not actually sent) because the user did not subscribe to the type of SMS

}；

While the present invention has been described with respect to the embodiments, those skilled in the art will appreciate that there are numerous variations and permutations of the present invention without departing from the spirit of the invention, and it is intended that the appended claims cover such variations and modifications as fall within the true spirit of the invention.

Claims

1. An apparatus for cluster system monitoring, comprising:

the user device is connected with the user wireless transceiver and used for receiving the monitoring data of the user wireless transceiver and displaying the monitoring data to a user;

the server includes:

the BSP device module of the basic service provider is used for realizing the monitoring data of each node machine in the collector group system;

and the database is used for realizing the monitoring data of the node machine acquired by the storage BSP module.

2. The apparatus of claim 1, further comprising local area network means, respectively connected to said server and said wireless network bridge connection means, for enabling connection of said server and said wireless network bridge connection means.

3. The apparatus of claim 1 or 2, wherein the server further comprises:

4. The apparatus according to claim 3, wherein the configuration means includes a short message reminding means for implementing an interface between the BSP and the wireless network bridge connection means, and filtering short message information customized by a cluster user; the wireless network bridge device comprises a controllable device and is used for sending the data packet of the short message information through a socket so as to be processed into the wirelessly sent short message by a common configuration program in the controllable device.

5. The apparatus of claim 2, wherein the local area network apparatus comprises: local area network: coupled to the server, the data transmission is realized based on a TCP/IP protocol; a network hub: a connection for managing a local area network; a network controller: the remote login management system is used for managing remote login management of the server through the local area network and the network hub.

6. The apparatus of claim 2, wherein said user wireless transceiving means comprises a PDA wireless network card for communicating a user's PDA with said wireless network bridge device; the user device is a PDA, and the PDA wireless network card is integrated in the PDA.

7. The apparatus of claim 6, wherein the server comprises: the data dividing device is used for classifying the collected monitoring data of each node according to the cabinet or the cluster; and the display processing device is used for processing the classified monitoring data into a data format displayed by a histogram.

8. A method of fleet system monitoring, comprising the steps of:

presetting a preset condition for a user to acquire monitoring data;

9. The method of claim 8, wherein the step of transmitting the filtered monitoring data to a user comprises: the screened monitoring data is transmitted to the user in a manner that the information is displayed in a bar graph supported by the PDA.

10. The method of claim 9, wherein the step of transmitting the monitoring data to a user comprises: after the webpage program of the server receives the message of requesting to send information of the user, the webpage program of the server generates query content according to the set data format and the preset condition, and transmits the query content to a service program so as to call an interface provided by a program supporting file uploading to query a database; then the inquired data is processed and recombined by the service program to generate a dynamic webpage; the server's web page program organizes the HTML language display interface for delivery to the user's PDA display.

11. The method of claim 8, wherein the step of transmitting the filtered monitoring data to a user comprises: transmitting the screened monitoring data to a user in a short message format; the predetermined conditions include fleet catastrophic alarm conditions and/or general alarm conditions and/or minor fault alarm conditions.

12. The method of claim 11, wherein,

the step of obtaining the monitoring data of each node of the cluster system comprises: the BSP module provided by the server collects monitoring data of each node machine through a local area network and a serial port;

the step of storing the acquired monitoring data comprises: writing the acquired monitoring data into a database through a first application program interface, and maintaining the monitoring data;

the step of transmitting the converted monitoring data to the user according to the set predetermined condition for the user to obtain the monitoring data comprises comparing the monitoring data with a predetermined threshold value in the predetermined condition, and if the predetermined threshold value is exceeded, generating an alarm event and transmitting the alarm event through a second application program interface.