[ 
https://issues.apache.org/jira/browse/HBASE-18625?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16296291#comment-16296291
 ] 

huaxiang sun edited comment on HBASE-18625 at 12/19/17 6:27 AM:
----------------------------------------------------------------

One more point, if the master crashes after the primary parent region split and 
right before handling replica parent region split, can the new master recover 
the state correctly? My current thinking is no, but I need some time to think 
through this case.


was (Author: huaxiang):
One more point, if the master crashes after the primary parent region split and 
right before handling replica parent region split, does the new master can 
recover the state correctly? My current thinking is no, but I need some time to 
think through this case.

> Splitting of region with replica, doesn't update region list in 
> serverHolding. A server crash leads to overlap.
> ---------------------------------------------------------------------------------------------------------------
>
>                 Key: HBASE-18625
>                 URL: https://issues.apache.org/jira/browse/HBASE-18625
>             Project: HBase
>          Issue Type: Bug
>    Affects Versions: 1.2.6
>            Reporter: Igloo
>            Assignee: huaxiang sun
>             Fix For: 1.2.8
>
>         Attachments: HBASE-18625-branch-1-v001.patch, 
> HBASE-18625-branch-1-v002.patch
>
>
> The situation can appear in following steps in release hbase1.2.6
> 1. create 'testtable', 'info', {REGION_REPLICATION=>2}
> 2. write somerecords into 'testtable'
> 3. split the table 'testtable'
> 4. after the spliting, the serverHoldings in RegionStates still holds the 
> regioninfo for the replica of parent region
> 5. restart the regionserver where the parent replica-region located
> 6. the offlined replica of parent region will be assigned in 
> ServerCrashProcedure. 
> hbase hbck 'testtable‘
> ERROR: Region { meta => null, hdfs => null, deployed => 
> qabb-qa-hdp-hbase1,16020,1503022958093;testtable,,1503022907686_0001.42d11cfe195b3cc4d08b2c078a687f6d
>     ., replicaId => 1 } not in META, but deployed on 
> qabb-qa-hdp-hbase1,16020,1503022958093
>  18 ERROR: No regioninfo in Meta or HDFS. { meta => null, hdfs => null, 
> deployed => 
> qabb-qa-hdp-hbase1,16020,1503022958093;testtable,,1503022907686_0001.42d11cfe 
>    195b3cc4d08b2c078a687f6d., replicaId => 1 }



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

Reply via email to