Determining a backup server for a session based on a deterministic mechanism and the session's key value

• US 8,090,829 B1
• Filed: 04/23/2004
• Issued: 01/03/2012
• Est. Priority Date: 04/23/2004
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Oracle America, Inc. (Oracle Corporation)
• Inventors: Galperin, Viatcheslav
• Primary Examiner(s): Nguyen, Thu
• Assistant Examiner(s): Greene, Joseph
• Application Number:
  US10/831,250
• Time in Patent Office: 2,811 Days
• Field of Search: 709/226, 709/240
• US Class Current: 709/226
• CPC Class Codes: G06F 11/2023 Failover techniques
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 11/2048 where the redundant components share neither address space nor persistent storage

Abstract

A mechanism is disclosed for determining, based on a deterministic mechanism and a session'"'"'s key value, a backup server for the session. In one embodiment, each session has a different key value, and each key value can be processed through the deterministic mechanism to give rise to a separate ordering of a result value set. Each server of a plurality of servers is associated with a different one of the set'"'"'s values. A request pertaining to a particular session is received. The request indicates the particular session'"'"'s key value. The deterministic mechanism can generate result values, in a particular order to which the session'"'"'s key value gives rise, until a result value associated with a currently operational server is generated. The request is sent to that server if the request'"'"'s primary server is not currently operational. A pseudo-random number generator seeded with the key value may generate the result values.

Claims

1. A machine-implemented method, comprising:

receiving, by a first backup server, a first request comprising a first session key value and a first primary server identifier, wherein the first backup server is one of a plurality of servers;

making a first determination, by the first backup server, that the first request is associated with a first primary server based on the first primary server identifier, wherein the first primary server is one of the plurality of servers;

making a second determination, by the first backup server, that the first primary server is not currently operational;

generating, by the first backup server based on the second determination, a first ordering of potential backup servers by applying a deterministic algorithm to the first session key value;

selecting, by the first backup server, a second backup server based on a first-ordered value from the first ordering of potential backup servers, wherein the second backup server is one of the plurality of servers;

sending, by the first backup server based on the selecting, the first request to the second backup server;

receiving, by the second backup server, the first request;

making a third determination, by the second backup server, that the first request is associated with the first primary server based on the first primary server identifier;

making a fourth determination, by the second backup server, that the first primary server is not currently operational;

generating, by the second backup server based on the fourth determination, the first ordering of potential backup servers by applying the deterministic algorithm to the first session key value;

obtaining the first-ordered value from the first ordering of potential backup servers; and

processing, by the second backup server, the first request based on a determination that the first-ordered value from the first ordering of potential backup servers is associated with the second backup server.

2. The method of claim 1, further comprising:

receiving, by the first backup server, a second request comprising a second session key value and a second primary server identifier;

making a fifth determination, by the first backup server, that the second request is associated with a second primary server based on the second primary server identifier, wherein the second primary server is one of the plurality of servers;

making a sixth determination, by the first backup server, that the second primary server is not currently operational;

generating, by the first backup server based on the sixth determination, a second ordering of potential backup servers by applying the deterministic algorithm to the second session key value;

obtaining, by the first backup server, a first-ordered value from the second ordering of potential backup servers; and

processing, by the first backup server, the second request based on a determination that the first-ordered value from the second ordering of potential backup servers is associated with the first backup server.

7. The method of claim 1, wherein the first primary server is a third backup server.

8. The method of claim 1, wherein making the first determination comprises:

making a third determination, by the first backup server, that session state information associated with the first session key value is not stored in a session table on the first backup server.

9. The method of claim 1, further comprising:

receiving, by the first backup server, a second request comprising a second session key value and a second primary server identifier;

making a fifth determination, by the first backup server, that the second request is associated with a second primary server based on the second primary server identifier, wherein the second primary server is one of the plurality of servers;

making a sixth determination, by the first backup server, that the second backup server is operational; and

sending the second request to the second backup server based on the sixth determination.

3. A non-transitory machine-readable medium, comprising instructions, which when executed by a processor perform a method, the method comprising:

receiving, by a first backup server, a first request comprising a first session key value and a first primary server identifier, wherein the first backup server is one of a plurality of servers;

making a first determination, by the first backup server, that the first request is associated with a first primary server based on the first primary server identifier, wherein the first primary server is one of the plurality of servers;

making a second determination, by the first backup server, that the first primary server is not currently operational;

generating, by the first backup server based on the second determination, a first ordering of potential backup servers by applying a deterministic algorithm to the first session key value;

selecting, by the first backup server, a second backup server based on a first-ordered value from the first ordering of potential backup servers, wherein the second backup server is one of the plurality of servers;

sending, by the first backup server based on the selecting, the first request to the second backup server;

receiving, by the second backup server, the first request;

making a third determination, by the second backup server, that the first request is associated with the first primary server based on the first primary server identifier;

making a fourth determination, by the second backup server, that the first primary server is not currently operational;

generating, by the second backup server based on the fourth determination, the first ordering of potential backup servers by applying the deterministic algorithm to the first session key value;

obtaining, the first-ordered value from the first ordering of potential backup servers; and

processing, by the second backup server, the first request based on a determination that the first-ordered value is associated with the second backup server.

4. The non-transitory machine-readable medium of claim 3, the method further comprising:

receiving, by the first backup server, a second request comprising a second session key value and a second primary server identifier;

making a fifth determination, by the first backup server, that the second request is associated with a second primary server based on the second primary server identifier, wherein the second primary server is one of the plurality of servers;

making a sixth determination, by the first backup server, that the second primary server is not currently operational;

generating, by the first backup server, a second ordering of potential backup servers by applying the deterministic algorithm to the second session key value;

obtaining, by the first backup server, a first-ordered value from the second ordering of potential backup servers; and

processing, by the first backup server, the second request based on a determination that the first-ordered value from the second ordering of potential backup servers is associated with the first backup server.

10. The non-transitory machine-readable medium of claim 3, wherein the first primary server is a third backup server.

11. The non-transitory machine-readable medium of claim 3, wherein making the first determination a first comprises:

making a third determination, by the first backup server, that session state information associated with the first session key value is not stored in a session table on the first backup server.

12. The non-transitory machine-readable medium of claim 3, the method further comprising:

receiving, by the first backup server, a second request comprising a second session key value and a second primary server identifier;

making a fifth determination, by the first backup server, that the second request is associated with a second primary server based on the second primary server identifier, wherein the second primary server is one of the plurality of servers;

making a sixth determination, by the first backup server, that the second backup server is operational; and

sending the second request to the second backup server based on the sixth determination.

5. An apparatus comprising:

a first backup server configured to;

receive a first request comprising a first session key value and a first primary server identifier;

make a first determination that the first request is associated with a first primary server based on the first primary server identifier, wherein the first primary server is one of a plurality of servers;

make a second determination that the first primary server is not currently operational;

generate a first ordering of potential backup servers by applying a deterministic algorithm to the first session key value;

select a second backup server based on a first-ordered value from the first ordering of potential backup servers;

send the first request to the second backup server;

the second backup server configured to;

receive the first request;

make a third determination that the first request is associated with the first primary server based on the first primary server identifier;

make a fourth determination that the first primary server is not currently operational;

generate, based on the fourth determination, the first ordering of potential backup servers by applying the deterministic algorithm to the first session key value;

obtain the first-ordered value from the first ordering of potential backup servers; and

process the first request based on a determination that the first-ordered value is associated with the second backup server.

6. The apparatus of claim 5, wherein the first backup server is further configured to;

receive a second request comprising a second session key value and a second primary server identifier;

make a fifth determination that the second request is associated with a second primary server based on the second primary server identifier, wherein the second primary server is one of the plurality of servers;

make a sixth determination that the second primary server is not currently operational;

based on the determination that the second primary server is not currently operational;

obtain the set of values;

seed the deterministic pseudo-random number generator with the second session key value;

generate, based on the sixth determination, a second ordering of potential backup servers by applying the deterministic algorithm to the second session key value;

obtain a first-ordered value from the second ordering of potential backup servers; and

process the second request based on a determination that the first-ordered value from the second ordering of potential backup servers is associated with the first backup server.

13. The apparatus of claim 5, wherein the first primary server is a third backup server.

14. The apparatus of claim 5, wherein making the first determination comprises:

making a third determination, by the first backup server, that session state information associated with the first session key value is not stored in a session table on the first backup server.

15. The apparatus of claim 5, wherein the first backup server is further configured to:

receive a second request comprising a second session key value and a second primary server identifier;

make a fifth determination, by the first backup server, that the second request is associated with a second primary server based on the second primary server identifier, wherein the second primary server is one of the plurality of servers;

make a sixth determination, that the second backup server is operational; and

send the second request to the second backup server based on the sixth determination.