[
https://issues.apache.org/jira/browse/HBASE-8028?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13596319#comment-13596319
]
Himanshu Vashishtha commented on HBASE-8028:
--------------------------------------------
Looked more into it.
For CF which don't have 1-version limitation, we can simply remove the added
kvs.
Its the upsert logic which makes this tricky: depending on the lowestreadpoint
at the time when it is doing the append operation, it decides whether it should
do an update (remove the existing kv), or insert a new kv. It doesn't keep a
history of its actions, though, which makes it difficult to decide the roll
back process.
[~lhofhansl]: What's your take on this? Is this the reason why Append doesn't
have a rollback scheme? :)
> Append, Increment don't handle wal-sync exceptions correctly
> ------------------------------------------------------------
>
> Key: HBASE-8028
> URL: https://issues.apache.org/jira/browse/HBASE-8028
> Project: HBase
> Issue Type: Bug
> Components: regionserver
> Affects Versions: 0.94.5
> Reporter: Himanshu Vashishtha
> Assignee: Himanshu Vashishtha
> Fix For: 0.95.0
>
>
> In case there is an exception while doing the log-sync, the memstore is not
> rollbacked, while the mvcc is _always_ forwarded to the writeentry created at
> the beginning of the operation. This may lead to scanners seeing results
> which are not synched to the fs.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
