US 7,321,992 B1
Reducing application downtime in a cluster using user-defined rules for proactive failover
Prabhakar Krishnamurthy Vellore, Irvine, Calif. (US); Mukund Hari Sharma, Los Alamitos, Calif. (US); and Peng Liu, Tustin, Calif. (US)
Assigned to Unisys Corporation, Blue Bell, Pa. (US)
Filed on Mar. 28, 2003, as Appl. No. 10/401,478.
Application 10/401478 is a continuation in part of application No. 10/156486, filed on May 28, 2002.
Int. Cl. G06F 11/00 (2006.01)
U.S. Cl. 714—47  [714/13] 27 Claims
OG exemplary drawing
 
1. A method comprising the operations of:
(a) monitoring an event log of a first server node to check for user-specified application events;
(b) detecting one of the user-specified application events, corresponding to an impending failure in an application running on the first server node; and
(c) executing a proactive failover process, in automatic response to the detected impending failure, to transfer the application to a second server node for continued execution, the second server node being connected to the first server node in a cluster;
wherein the proactive failover process includes:
determining if the application is in off-line state; and
if the application is in off-line state:
logging the user-specified application event and corresponding action result; and
terminating the proactive failover process.