[
https://issues.apache.org/jira/browse/HBASE-21073?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16647013#comment-16647013
]
stack commented on HBASE-21073:
-------------------------------
What does this do (smile)?
There is already a log message (?) about master being in maintenance mode...
You saw that (Perhaps related to HBASE-16008 ?)? Has that been appropriated by
this patch?
bq. I have no idea what happens if we try to scan one of the user space tables.
A client would hang, right, then fail because it would try to go to server
hosting the user-space region. Should maintenance mode bring up the Master but
only have it listening for Connection from localhost?
bq. Why do I need to explicitly set TABLES_ON_MASTER to true in the test? I
haven't yet found where Master reads that value to know that it should report
as an RS.
hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java:
boolean tablesOnMaster = LoadBalancer.isTablesOnMaster(conf);
Is that it? This bit of code here is strange. In usual case -- no regions on
master -- we actually seem to hang here... which doesn't seem right.
Rather than fold this into HBASE-16008, HBASE-16008 should be purged or folded
in here. HBASE-16008's notion of 'maintenance mode' is strange... its setting a
running master into a 'state' so hbck1 could run.... and then restoring old
state after.
> "Maintenance mode" master
> -------------------------
>
> Key: HBASE-21073
> URL: https://issues.apache.org/jira/browse/HBASE-21073
> Project: HBase
> Issue Type: Sub-task
> Components: amv2, hbck2, master
> Reporter: stack
> Assignee: Mike Drob
> Priority: Major
> Attachments: HBASE-21073.master.001.patch,
> HBASE-21073.master.002.patch
>
>
> Make it so we can bring up a Master in "maintenance mode". This is parse of
> master wal procs but not taking on regionservers. It would be in a state
> where "repair" Procedures could run; e.g. a Procedure that could recover meta
> by looking for meta WALs, splitting them, dropping recovered.edits, and even
> making it so meta is readable. See parent issue for why needed (disaster
> recovery).
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
