[
https://issues.apache.org/jira/browse/HBASE-12035?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14314469#comment-14314469
]
Andrey Stepachev commented on HBASE-12035:
------------------------------------------
[~stack] as for migration, I've checked, it seems that it should work.
first TableStateManager will gather states from descriptors. If hbase is older
and have no data in descriptors,
it seems all table states become enabled.
then ZKDataMigrator will read states from zk and for old versions will update
tables to correct states using
new TableStateManager (which in turn set correct states in meta).
in case of newer version ZkDataMigrator shouldn't find anything valuable and
states will be those, which
was got from descriptors (or simply defaulted to ENABLED if there was no states
at all).
seems mind replay shows that should work.
> Keep table state in META
> ------------------------
>
> Key: HBASE-12035
> URL: https://issues.apache.org/jira/browse/HBASE-12035
> Project: HBase
> Issue Type: Improvement
> Components: Client, master
> Affects Versions: 2.0.0
> Reporter: Enis Soztutar
> Assignee: Andrey Stepachev
> Priority: Critical
> Fix For: 2.0.0
>
> Attachments: 12035v2.txt, HBASE-12035 (1) (1).patch, HBASE-12035 (1)
> (1).patch, HBASE-12035 (1).patch, HBASE-12035 (2).patch, HBASE-12035
> (2).patch, HBASE-12035.patch, HBASE-12035.patch, HBASE-12035.patch,
> HBASE-12035.patch, HBASE-12035.patch, HBASE-12035.patch, HBASE-12035.patch,
> HBASE-12035.patch, HBASE-12035.patch, HBASE-12035.patch, HBASE-12035.patch,
> HBASE-12035.patch, HBASE-12035.patch, HBASE-12035.patch, HBASE-12035.patch,
> HBASE-12035.patch, HBASE-12035.patch
>
>
> HBASE-7767 moved table enabled|disabled state to be kept in hdfs instead of
> zookeeper. isTableDisabled() which is used in
> HConnectionImplementation.relocateRegion() now became a master RPC call
> rather than a zookeeper client call. Since we do relocateRegion() calls
> everytime we want to relocate a region (region moved, RS down, etc) this
> implies that when the master is down, the some of the clients for uncached
> regions will be affected.
> See HBASE-7767 and HBASE-11974 for some more background.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
