[
https://issues.apache.org/jira/browse/IGNITE-5813?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16147123#comment-16147123
]
Andrey Gura commented on IGNITE-5813:
-------------------------------------
We can't just invalidate entries because in this case store and cache states
will be still inconsistent. It's ok for get operations with readThrough enabled
but SQl queries will return wrong results.
> Inconsistent cache store state when originating node fails on commit.
> ---------------------------------------------------------------------
>
> Key: IGNITE-5813
> URL: https://issues.apache.org/jira/browse/IGNITE-5813
> Project: Ignite
> Issue Type: Bug
> Components: cache
> Reporter: Andrew Mashenkov
> Assignee: Andrew Mashenkov
> Fix For: 2.3
>
> Attachments: IgniteTxRecoveryAfterStoreCommitSelfTest.java
>
>
> Same tx can be rolled back by cache and commited by CacheStore.
> It is possible when originating tx node commit tx successfully, but fails to
> notify other nodes.
> PFA reproducer.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
