hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jie Huang (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-6380) bulkload should update the store.storeSize
Date Thu, 12 Jul 2012 12:07:34 GMT

    [ https://issues.apache.org/jira/browse/HBASE-6380?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13412725#comment-13412725

Jie Huang commented on HBASE-6380:

We found that action of force-split on certain table (after bulkloading) failed, and we can
see following debug log
DEBUG org.apache.hadoop.hbase.regionserver.CompactSplitThread: Region t1,,1342088772367.767829f873354d78a742084c22f47895.
not splittable because midkey=null
But actually we can get the non-empty midkey value in getSplitPoint() function. The only reason
is that the storeSize and largestStoreSize are zero, which makes the splitPointFromLargetStore
returned with NULL (its initial value). 
byte[] splitPointFromLargestStore = null;
    long largestStoreSize = 0;
    for (Store s : stores.values()) {
      byte[] splitPoint = s.getSplitPoint();
      long storeSize = s.getSize();
      if (splitPoint != null && largestStoreSize < storeSize) {
        splitPointFromLargestStore = splitPoint;
        largestStoreSize = storeSize;
    return splitPointFromLargestStore;
The simple solution is to update the store's size in the bulkload() function. After applied
attached patch, the problem is solved. 
> bulkload should update the store.storeSize
> ------------------------------------------
>                 Key: HBASE-6380
>                 URL: https://issues.apache.org/jira/browse/HBASE-6380
>             Project: HBase
>          Issue Type: Bug
>          Components: regionserver
>    Affects Versions: 0.94.0, 0.96.0
>            Reporter: Jie Huang
>            Priority: Critical
>         Attachments: hbase-6380_0_94_0.patch
> After bulkloading some HFiles into the Table, we found the force-split didn't work because
of the MidKey == NULL. Only if we re-booted the HBase service, the force-split can work normally.

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


View raw message