System and method for continuous monitoring and measurement of performance of computers on network

• US 7,577,701 B1
• Filed: 01/22/2001
• Issued: 08/18/2009
• Est. Priority Date: 01/22/2001
• Status: Expired due to Fees

• Current Assignee: Magnafox LLC
• Original Assignee: InsightETE Corporation
• Inventors: Johns, William M., Lindau, Brian G.
• Primary Examiner(s): Lin; Kenny S
• Assistant Examiner(s): TRUONG, LAN DAI T
• Application Number:
  US09/765,431
• Time in Patent Office: 3,130 Days
• Field of Search: 709/224, 709/206, 709/203, 709/204, 709/223, 709/217, 709/220, 790/224
• US Class Current: 709/203
• CPC Class Codes: H04L 41/0213 Standardised network management protocols, e.g. simple network management protocol [SNMP]
• CPC Class Codes: H04L 41/0631 using root cause analysis; using analysis of correlation between notifications, alarms or events based on decision criteria, e.g. hierarchy, tree or time analysis
• CPC Class Codes: H04L 41/5009 Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
• CPC Class Codes: H04L 43/00 Arrangements for monitoring or testing data switching networks
• CPC Class Codes: H04L 43/028 by filtering
• CPC Class Codes: H04L 43/0817 by checking functioning
• CPC Class Codes: H04L 43/106 using time related information in packets, e.g. by adding timestamps
• CPC Class Codes: H04L 43/12 Network monitoring probes

Abstract

Performance on a networked system is monitored through five interlocking monitoring techniques. A network monitoring process causes pseudo messages to be sent from various points in the network and monitors the progress of the pseudo messages. A client-server monitoring process monitors server availability and receives logs from various client-server processes and determines the performance of both server and server processes, which can be grouped by end-user location. A mainframe monitoring process monitors the performance of the various hardware and software components on a mainframe. A reporting and administration process compiles reports and makes them remotely available over the Internet. An e-business monitoring process monitors servers accessible to customers over the Internet by use of a Java applet which sends time stamps indicating access. The five processes are tied together through a central data repository, called a data mart.

Claims

1. A method for monitoring performance and availability of application servers on a network, including a percentage of time that each of the application servers is available to an end user relative to the time the application servers are intended to be available and a responsiveness of the application servers to the end user in terms of a delay between the end user'"'"'s entering data into a workstation keyboard and a response from one of the application servers with new data on the user'"'"'s workstation screen, the method comprising:

(a) running at least one performance monitor process on the network, said at least one performance monitor process watching network activity to and from the application servers to entry servers which connect the network to the end user'"'"'s workstation and creating a transaction response time log and activity audit trail for the network;

(b) running a network monitor manager process on the network, for consolidating information from the transaction response time log;

(c) establishing a connection from the network monitor manager process to said at least one performance monitor process to control said at least one performance monitor to send a pseudo message for tracking time in the network to an entry server to determine said network availability; and

(d) receiving the pseudo message from said at least one performance monitor process and determining a response for the pseudo message for each segment of the network traversed by the pseudo message to determine where problems regarding said availability exist within the network connection for the entry server;

(i) running a client-server monitoring process on a server dedicated to the client-server monitoring process;

(j) receiving, in the client-server monitoring process, information about transactions executed by production applications on the network; and

(k) determining performance and availability of the production applications in accordance with the information received in step (j);

wherein step (j) comprises running a filtering agent on each or on behalf of each of the production applications to convert the information from application logs into a form usable by the client-server monitoring process;

wherein the network comprises a mainframe having at least one logical partition which generates an application log; and

the method further comprises step (l) monitoring the application log through a mainframe monitoring process.

2. The method of claim 1, further comprising:

(e) running at least one availability monitor process on the network;

(f) from the response determined in step (d), detecting at least one possibly failed component of the network;

(g) sending a message from the at least one availability monitor process to the at least one possibly failed component; and

(h) determining, in accordance with a result of the message, whether the at least one possibly failed component has failed.

3. The method of claim 1, wherein:

the application log comprises transaction entries having end-user addresses; and

step (l) comprises categorizing the transaction entries by the end-user addresses.

4. The method of claim 1, wherein each said filtering agent detects processes running on the network and cross-references the detected processes to known processes, and further comprising forming an event correlation engine in accordance with the detected processes.

5. The method of claim 3, further comprising (m) generating a performance report for the network through an administrative process and making the report available over a data network.

6. The method of claim 4, wherein each said filtering agent detects changes to the processes running on the network, and further comprising maintaining the event correlation engine in accordance with the detected changes to the processes.

7. The method of claim 5, wherein the data network comprises the Internet.

8. The method of claim 6, further comprising, when it is determined in step (k) that the performance or the availability of one of the production applications is impaired, determining and reporting a cause of impairment and its corresponding effect on a service level agreement (SLA) in accordance with the event correlation engine.

9. The method of claim 7, further comprising:

(n) receiving, in the client-server monitoring process, information about transactions executed by e-commerce applications on the network; and

(o) determining performance and availability of the e-commerce applications in accordance with the information received in step (n) through an e-commerce monitoring process.

10. The method of claim 9, wherein at least one of the e-commerce applications makes at least one Web page accessible to customers, and wherein step (n) comprises placing code in the at least one Web page, the code sending time stamps to the client-server monitoring process when the code is accessed.

11. The method of claim 10, further comprising providing a central data repository, and wherein the network monitor manager process, the client-server monitoring process, the mainframe monitoring process, the administrative process, and the e-commerce monitoring process communicate with one another through the central data repository.