[
https://issues.apache.org/jira/browse/HUDI-1580?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17311187#comment-17311187
]
Nishith Agarwal commented on HUDI-1580:
---------------------------------------
[~shivnarayan] [~ryanpife] and I have chatted about this before which resulted
in this ticket. There are mechanisms at Uber but unfortunately that is not
upstream-able.
[~ryanpife] I know you guys were working on some aspects of this here, any
progress ?
> Disaster Recovery in case where HBASE index table becomes unrecoverable
> -----------------------------------------------------------------------
>
> Key: HUDI-1580
> URL: https://issues.apache.org/jira/browse/HUDI-1580
> Project: Apache Hudi
> Issue Type: New Feature
> Components: Index
> Affects Versions: 0.9.0
> Reporter: Ryan Pifer
> Priority: Major
> Labels: sev:high, user-support-issues
> Fix For: 0.9.0
>
>
> If an HBASE table becomes corrupted for any reason, we should think of a way
> that we can provide some sort of disaster recovery built in to re-build
> index. Currently only way to do this without any extra utilities is to
> rewrite entire dataset.
> Can we create a CLI command which rebuilds just index?
> Can we add checkpointing to hbase table to reduce disaster recovery time?
> i.e. only re-build index for records written after last checkpoint time
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
