Method and apparatus for collecting data for characterizing HTTP session workloads

• US 20070150600A1
• Filed: 12/22/2005
• Published: 06/28/2007
• Est. Priority Date: 12/22/2005
• Status: Abandoned Application

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Stecher, John, Santosuosso, John, Barsness, Eric
• Application Number:
  US11/316,285
• Publication Number:
  US 20070150600A1
• Time in Patent Office: Days
• US Class Current: 709/227
• CPC Class Codes: G06F 11/2097 maintaining the standby controller/processing unit updated initialisation or re-synchronisation thereof G06F11/1658 and subgroups
• CPC Class Codes: G06F 11/3447 Performance evaluation by modeling
• CPC Class Codes: G06F 11/3495 for systems
• CPC Class Codes: G06F 2201/875 Monitoring of systems including the internet
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1027 Persistence of sessions during load balancing
• CPC Class Codes: H04L 67/1034 Reaction to server failures by a load balancer
• CPC Class Codes: H04L 67/142 Managing session states for stateless protocols; Signalling session states; State transitions; Keeping-state mechanisms

Abstract

A web server comprising multiple computer systems includes at least one system which maintains a database of redundant session-persistent data on behalf of other systems which process HTTP session workloads for clients. A session analyzer extracts information from this redundancy database to characterize HTTP session workloads. Preferably, session workloads are processed by middleware comprising an application server and one or more applications. The application server maintains session-persistent data as data objects, and automatically updates the database of redundant data which responses are transmitted to a clients. The session analyzer independently extracts data from the redundant database to characterize the session workloads. Optionally, the analyzer includes an intelligent inferencing function which can make tuning recommendations from the extracted data.

Claims

1. A method for characterizing HTTP session workload in a web server, comprising the computer-executed steps of:

processing a plurality of HTTP sessions in a first computer system,

maintaining session-persistent state data in said first computer system for said plurality of HTTP sessions;

transmitting redundant session-persistent state data from said first computer system to a second computer system;

storing said redundant session-persistent state data in said second computer system, said second computer system supporting recovery of session state of said plurality of HTTP sessions using said redundant session-persistent state data stored in said second computer system; and

characterizing said HTTP session workload using said redundant session-persistent state data stored in said second computer system.

2. The method for characterizing HTTP session workload of claim 1, wherein said step of storing said redundant session-persistent state data in said second computer system comprises storing said redundant session-persistent state data in a structured database in said second computer system, said structured database having at least one table having a plurality of entries, each entry corresponding to a respective HTTP session of said plurality of HTTP sessions.

3. The method for characterizing HTTP session workload of claim 2, wherein said step of characterizing said HTTP session workload using said redundant session-persistent state data comprises executing one or more queries against said structured database to retrieve data stored therein.

4. The method for characterizing HTTP session workload of claim 2, wherein said step of characterizing HTTP session workload using said redundant session state data comprises monitoring changes to said structured database, and using historical data concerning changes to said structured database to characterize said HTTP session workload.

5. The method for characterizing HTTP session workload of claim 2, wherein said step of characterizing HTTP session workload using said redundant session-persistent state data comprises using data from said structured database to draw inferences with respect to tuning parameters of said first computer system, said inferences being automatically drawn by a computer-executed process.

6. The method for characterizing HTTP session workload of claim 2, wherein said step of characterizing HTTP session workload using said redundant session-persistent state data comprises using data from said structured database to derive at least one characterization parameter from the set of characterization parameters consisting of:

(a) a number of database update operations measured over at least one interval;

(b) an average update operation size;

(c) an average size of said entries;

(d) an overall size of said structured database;

(e) an average column size for at least one column of said structured database;

(f) data types being used by entries of said structured database;

(g) an average session lifespan;

(h) an average response time for updates to said database;

(i) a number of writes to a log of said database; and

(j) an average wait time to disk in said second computer system.

7. The method for characterizing HTTP session workload of claim 1, wherein said first computer system comprises an application server supporting a session server application for processing HTTP sessions.

8. The method for characterizing HTTP session workload of claim 1, wherein said second computer system stores redundant session-persistent state data received from a plurality of first computer systems, each first computer system processing a respective plurality of HTTP sessions and maintaining session-persistent state data for the respective plurality of HTTP sessions.

9. The method for characterizing HTTP session workload of claim 1, wherein said first computer system and said second computer system are coupled to a common local area network, and wherein said step of transmitting redundant session-persistent state data from said first computer system to said second computer system comprises transmitting said redundant session-persistent state data across said local area network.

10. A computer program product for characterizing HTTP session workload of a web server, comprising:

a plurality of computer-executable instructions recorded on signal-bearing media, wherein said instructions, when executed by at least one computer system, cause the at least one computer system to perform the steps of;

accessing redundant session-persistent state data stored in a first computer system, said redundant session-persistent data being received in said first computer system from a second computer system, said redundant session-persistent state data representing session-persistent state maintained by said second computer system for processing a plurality of HTTP sessions by said second computer system, said redundant session-persistent sate data being stored in said first computer system in a manner supporting recovery of session state of said plurality of HTTP sessions using said redundant session-persistent state data; and

characterizing said HTTP session workload using said redundant session-persistent state data accessed by said accessing step.

11. The computer program product of claim 10, wherein said redundant session-persistent state data stored in said first computer system comprises redundant session-persistent state data stored in a structured database in said first computer system, said structured database having at least one table having a plurality of entries, each entry corresponding to a respective HTTP session of said plurality of HTTP sessions.

12. The computer program product of claim 11, wherein said step of characterizing said HTTP session workload using said redundant session-persistent state data comprises executing one or more queries against said structured database to retrieve data stored therein.

13. The computer program product of claim 11, wherein said step of characterizing HTTP session workload using said redundant session state data comprises monitoring changes to said structured database, and using historical data concerning changes to said structured database to characterize said HTTP session workload.

14. The computer program product of claim 11, wherein said step of characterizing HTTP session workload using said redundant session-persistent state data comprises using data from said structured database to draw inferences with respect to tuning parameters of said second computer system, said inferences being automatically drawn by a computer-executed process.

15. The computer program product of claim 11, wherein said step of characterizing HTTP session workload using said redundant session-persistent state data comprises using data from said structured database to derive at least one characterization parameter from the set of characterization parameters consisting of:

(a) a number of database update operations measured over at least one interval;

(b) an average update operation size;

(c) an average size of said entries;

(d) an overall size of said structured database;

(e) an average column size for at least one column of said structured database;

(f) data types being used by entries of said structured database;

(g) an average session lifespan;

(h) an average response time for updates to said database;

(i) a number of writes to a log of said database; and

(j) an average wait time to disk in said first computer system.

16. An apparatus for characterizing HTTP session workloads, comprising:

a session redundancy computer system having at least one processor and a data storage for storing redundant session-persistent data received from at least one session server computer system external to said session redundancy computer system, said at least one session server computer system processing a plurality of HTTP sessions and maintaining session-persistent state data for said plurality of HTTP sessions;

wherein said session redundancy computer system supports recovery of session state of said plurality of HTTP sessions using said redundant session-persistent state data stored in said data storage; and

an HTTP session analyzer function embodied as a plurality of instructions executable on a computer system, said analyzer function characterizing workload of said plurality of HTTP sessions on said at least one server computer system using said redundant session-persistent state data stored in said data storage of said session redundancy computer system.

17. The apparatus of claim 16, wherein said HTTP session analyzer function resides in said session redundancy computer system and executes on said at least one processor of said session redundancy computer system.

18. The apparatus of claim 16, wherein said redundant session-persistent state data stored in said session redundancy computer system comprises redundant session-persistent state data stored in a structured database in said session redundancy computer system, said structured database having at least one table having a plurality of entries, each entry corresponding to a respective HTTP session of said plurality of HTTP sessions.

19. The apparatus of claim 16, further comprising a plurality of session server computer systems, each session server computer system processing a respective plurality of HTTP session and maintaining said session-persistent state data for each respective HTTP session of the respective said plurality of HTTP sessions.

20. The apparatus of claim 19, further comprising at least one local area network providing a communications path for transmitting said redundant session-persistent state data from each said session server computer system to said session redundancy computer system.