In-line server health checking

• US 7,313,735 B1
• Filed: 04/30/2004
• Issued: 12/25/2007
• Est. Priority Date: 04/30/2004
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Caplan, David S., Levergood, Thomas M., Holland, Peter R.
• Primary Examiner(s): Beausoliel, Robert
• Assistant Examiner(s): Riad, Amine
• Application Number:
  US10/836,676
• Time in Patent Office: 1,334 Days
• Field of Search: 714/47, 709/223, 709/224, 709/225, 709/226
• US Class Current: 714/47.3
• CPC Class Codes: G06F 11/2294 by remote test
• CPC Class Codes: H04L 43/0817 by checking functioning
• CPC Class Codes: H04L 43/10 Active monitoring, e.g. heartbeat, ping or trace-route

Abstract

A system and method provide in-line server health checks to determine server health without having to wait for a next out-of-band polling interval as with conventional methods. By determining server health directly resulting from TCP and HTTP client requests, multiple connection retries on an unavailable server are prevented. The method for providing server health checking of a server in a network includes activating an in-line server health checking mechanism by monitoring a data path of a server as to health status. Results from the health status analysis are ultimately reported to an authority on server health.

Claims

1. In an electronic device, a health check system, comprising:

an in-line server health check mechanism, comprising;

an in-line health analysis mechanism for monitoring a server in a data path of the server relative to selected criteria to determine health status; and

a reporting mechanism for conveying the health status of the server;

a selection mechanism, for tracking the server health status; and

an out-of-band server health check mechanism for generating polls of the server outside of the data path, the out-of-band server health check mechanism further comprising a return-to-health mechanism for polling a failed server at repeated intervals to determine if the server is available.

2. The health check system of claim 1, wherein the health status is used in the determination of proportional share of future server load.

3. The health check system of claim 1, wherein the in-line server health check mechanism further comprises a communication link with an out-of-band server health check mechanism.

4. The method of claim 1 wherein the in-line server health check mechanism determines server health directly from the data path such that multiple connection retries on an unavailable server are avoided.

5. The method of claim 4 wherein the data path includes TCP and HTTP client requests.

6. In a network supporting a network protocol, a method of checking a health status of a server, comprising:

activating an in-line server health checking mechanism by monitoring a data path of the server;

analyzing a response from the server relative to selected criteria to determine health status;

reporting the health status of the server; and

tracking the server health status, tracking further comprising;

initiating polling via an the out-of-band server to determine the health status of the server; and

communicating the server health status to the out-of-band server health check mechanism, the a health status from the out-of-band server health check mechanism having priority over a health status from the in-line server health check mechanism according to predetermined parameters.

7. The method of claim 6, wherein tracking comprises maintaining a table of health status information.

8. The method of claim 6, wherein reporting the health status of the server comprises reporting that the server is one of up, downgraded, and down.

9. The method of claim 8, further comprising initiating a return-to-health query if the server is down.

10. The method of claim 9, wherein the return-to-health query repeats a selected number of times at a selected interval.

11. The method of claim 6 wherein the in-line server health check mechanism determines server health directly from the data path such that multiple connection retries on an unavailable server are avoided.

12. A medium for use in a network environment on an electronic device, the medium holding instructions executable using the electronic device for performing a method of checking a health status of a server, the method comprising:

activating an in-line server health checking mechanism by monitoring a data path of the server;

analyzing a response from the server relative to selected criteria to determine health status;

reporting the health status of the server, reporting the health status of the server further comprising reporting that the server is one of up, downgraded, and down;

initiating a return-to-health query if the server is down; and

tracking the server health status.

13. The medium of claim 12, wherein tracking comprises maintaining a table of health status information.

14. The medium of claim 12, further comprising communicating the server health status of an out-of-band server health check mechanism.

15. The medium of claim 14, further comprising the out-of-band server health mechanism initiating polling to determine the health status of the server.

16. The medium of claim 15, wherein a health status from the out-of-band server health check mechanism has priority over a health status from the in-line server health check mechanism according to predetermined parameters.

17. The medium of claim 12, wherein the return-to-health query repeats a selected number of times at a selected interval.

18. The method of claim 12 wherein the in-line server health check mechanism determines server health directly from the data path such that multiple connection retries on an unavailable server are avoided.

19. The method of claim 18 wherein the data path includes TCP and HTTP client requests.