[
https://issues.apache.org/jira/browse/HBASE-20242?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Duo Zhang updated HBASE-20242:
------------------------------
Resolution: Fixed
Hadoop Flags: Reviewed
Release Note: Now when opening a region, we will store the current max
sequence id of the region to its max sequence id file instead of the 'next
sequence id'. This could avoid the sequence id bumping when we fail to open a
region, and also align to the behavior when we close a region.
Status: Resolved (was: Patch Available)
Pushed to master.
Thanks [~openinx] for reviewing.
> The open sequence number will grow if we fail to open a region after writing
> the max sequence id file
> -----------------------------------------------------------------------------------------------------
>
> Key: HBASE-20242
> URL: https://issues.apache.org/jira/browse/HBASE-20242
> Project: HBase
> Issue Type: Sub-task
> Reporter: Duo Zhang
> Assignee: Duo Zhang
> Priority: Major
> Fix For: 3.0.0
>
> Attachments: HBASE-20242.patch
>
>
> We write the current max sequence id plus one as the new max sequence id so
> if we fail before report RIT to master, when assign next time the open
> sequence number will increase by one.
> But it is possible that we fail before writing the open region marker so
> there will no actual WAL with the sequence id which is the open sequence
> number minus one. This will cause the serial replication to be stuck.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
