[
https://issues.apache.org/jira/browse/HBASE-19144?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16227729#comment-16227729
]
Andrew Purtell edited comment on HBASE-19144 at 10/31/17 10:46 PM:
-------------------------------------------------------------------
[~stack] reported the same thing with branch-2 on HBASE-17580.
Seems the decision we made to make FAILED_OPEN state something an operator has
to resolve by hand, or by hbck (someday), or by master failover, is a nasty
operational wrinkle when using RSgroups, because the possibility of all servers
in a RSGroup going down is higher than for the whole cluster.
I wonder if we can more generally consider kicking FAILED_OPEN state
assignments whenever a new server shows up, with hysteresis such that if a few
servers show up in a short period of time we don't pile assignments all on the
first one to show up.
was (Author: apurtell):
[~stack] reported the same thing with branch-2 on HBASE-17580.
Seems the decision we made to make FAILED_OPEN state something an operator has
to resolve by hand, or by hbck (someday), or by master failover, is a nasty
operational wrinkle when using RSgroups, because the possibility of all servers
in a RSGroup going down is higher than for the whole cluster.
> [RSgroups] Regions assigned to a RSGroup all go to FAILED_OPEN state when all
> servers in the group are unavailable
> ------------------------------------------------------------------------------------------------------------------
>
> Key: HBASE-19144
> URL: https://issues.apache.org/jira/browse/HBASE-19144
> Project: HBase
> Issue Type: Bug
> Affects Versions: 2.0.0, 3.0.0, 1.4.0, 1.5.0
> Reporter: Andrew Purtell
>
> After all servers in the RSgroup are down the regions cannot be opened
> anywhere and transition rapidly into FAILED_OPEN state.
>
> 2017-10-31 21:06:25,449 INFO [ProcedureExecutor-13] master.RegionStates:
> Transition {c6c8150c9f4b8df25ba32073f25a5143 state=OFFLINE, ts=1509483985448,
> server=node-5.cluster,16020,1509482700768} to
> {c6c8150c9f4b8df25ba32073f25a5143 state=FAILED_OPEN, ts=1509483985449,
> server=node-5.cluster,16020,1509482700768}
> 2017-10-31 21:06:25,449 WARN [ProcedureExecutor-13] master.RegionStates:
> Failed to open/close d4e2f173e31ffad6aac140f4bd7b02bc on
> node-5.cluster,16020,1509482700768, set to FAILED_OPEN
>
> Any region in FAILED_OPEN state has to be manually reassigned, or the master
> can be restarted and this will also cause reattempt of assignment of any
> regions in FAILED_OPEN state. This is not unexpected but is an operational
> headache. It would be better if the RSGroupInfoManager could automatically
> kick reassignments of regions in FAILED_OPEN state when servers rejoin the
> cluster.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
