How to Handle Store Failures Using Write-Behind Caching (Doc ID 978862.1)

Last updated on MARCH 12, 2024

Applies to:

Oracle Coherence - Version 3.5.0 to 3.5.3 [Release AS10g]
Information in this document applies to any platform.

Goal

The following how to shows how to deal with store failures in a write-behind scenario. The client performing the write to the cache will not have an exception thrown to it upon failure, so it is incapable of handling the exception. Here we setup a demo showing how a store failure will be re-queued and continually re-tried until it has successfully been stored.

Note that the behavior of the write-behind queueing changes in Coherence 3.6 and later, where the <write-requeue-threshold> no longer sets a maximum size for re-queued and simply determines how frequently information on the size of the queue is reported.