[ 
https://issues.apache.org/jira/browse/HBASE-5677?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13242118#comment-13242118
 ] 

xufeng commented on HBASE-5677:
-------------------------------

If region be assigned When the master is doing initialization(before do 
processFailover),the region will be duplicate openhandled.
because the unassigned node in zookeeper will be handled again in 
AssignmentManager#processFailover()

I use the 0.90 vsersion.
I found this issue in my cluster.

1.The system did not do balance:
{noformat}
Not running balancer because 2 region(s) in transition: 
{f4ff609df50e5bc9049fe202bb90f22e=hbase0205test,0038613850505050,1333033465665.f4ff609df50e5bc9049fe202bb90f22e.
 
state=OPEN, ts=1333036748502, 
febe5bb42ec841f7a9086d3b7bf0637c=hbase0205test,0038613802020202,1333033465474.febe5bb42ec841f7a9086d3b7bf0637c...
{noformat}

2.Choose f4ff609df50e5bc9049fe202bb90f22e as a simple to track.

3.In master log I found:
logA:
{noformat}
Line 17884: [2012-03-29 15:05:08,082] [DEBUG] 
[MASTER_OPEN_REGION-158-1-130-18:20000-1] 
[org.apache.hadoop.hbase.master.handler.OpenedRegionHandler 138] The master has 
opened the region 
hbase0205test,0038613850505050,1333033465665.f4ff609df50e5bc9049fe202bb90f22e. 
that was online on serverName=158-1-130-18,20020,1332952904731, 
load=(requests=4444, regions=728, usedHeap=141, maxHeap=8165)
{noformat}

logB:
{noformat}
=Line 17885: [2012-03-29 15:05:08,082] [DEBUG] [master-158-1-130-18:20000] 
[org.apache.hadoop.hbase.master.handler.OpenedRegionHandler 138] Handling 
OPENED event for 
hbase0205test,0038613850505050,1333033465665.f4ff609df50e5bc9049fe202bb90f22e. 
from serverName=158-1-130-18,20020,1332952904731, load=(requests=245, 
regions=758, usedHeap=145, maxHeap=8165); deleting unassigned node
Line 17897: [2012-03-29 15:05:08,084] [DEBUG] [master-158-1-130-18:20000] 
[org.apache.hadoop.hbase.zookeeper.ZKAssign 511] master:20000-0x236552a09e20353 
Deleting existing unassigned node for f4ff609df50e5bc9049fe202bb90f22e that is 
in expected state RS_ZK_REGION_OPENED
Line 17898: [2012-03-29 15:05:08,092] [WARN ] [master-158-1-130-18:20000] 
[org.apache.hadoop.hbase.master.handler.OpenedRegionHandler 123] The znode of 
the region 
hbase0205test,0038613850505050,1333033465665.f4ff609df50e5bc9049fe202bb90f22e. 
would have already been deleted
Line 17899: [2012-03-29 15:05:08,092] [ERROR] [master-158-1-130-18:20000] 
[org.apache.hadoop.hbase.master.handler.OpenedRegionHandler 97] The znode of 
region 
hbase0205test,0038613850505050,1333033465665.f4ff609df50e5bc9049fe202bb90f22e. 
could not be deleted.
{noformat}

4.The logA and logB should not appear at the same time,because belong to the 
same code in the region open flow.

5.So I ensure that this region has been handled duplicate.

6.Those log can explain what I write in Description:
Enable the table:
{noformat}
Line 16925: [2012-03-29 15:04:59,875] [DEBUG] 
[158-1-130-18:20000-org.apache.hadoop.hbase.master.handler.EnableTableHandler$BulkEnabler-0]
 [org.apache.hadoop.hbase.zookeeper.ZKAssign 289] 
master:20000-0x236552a09e20353 Creating (or updating) unassigned node for 
f4ff609df50e5bc9049fe202bb90f22e with OFFLINE state
{noformat}

Failover:
{noformat}
[2012-03-29 15:05:00,906] [INFO ] [master-158-1-130-18:20000] 
[org.apache.hadoop.hbase.master.AssignmentManager 284] Failed-over master needs 
to process 66 regions in transition
{noformat}
                
> The master never does balance because duplicate openhandled the one region
> --------------------------------------------------------------------------
>
>                 Key: HBASE-5677
>                 URL: https://issues.apache.org/jira/browse/HBASE-5677
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 0.90.6
>         Environment: 0.90
>            Reporter: xufeng
>            Assignee: xufeng
>
> If region be assigned When the master is doing initialization(before do 
> processFailover),the region will be duplicate openhandled.
> it cause the region in RIT,thus the master never does balance.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: 
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira


Reply via email to