[ 
https://issues.apache.org/jira/browse/HBASE-26249?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Xiaolin Ha updated HBASE-26249:
-------------------------------
    Description: 
HBASE-25213 introduced to trigger compact after bulkload, but if a store needs 
compaction, each file loaded to this store will add a compaction runner of this 
store to the compaction queue. Then repeatedly compute for needCompaction and 
selectFiles for this store, and the compaction of the store may be continuous.

We can mark the queued compaction runners for the stores, and avoid repeatedly 
adding same store compactions to the queue after bulk load files. This method 
can reduce the pressure during the period just after bulk loading and spread 
the pressure to other times by the periodic CompactionChecker.

 

  was:
HBASE-25213 introduced to compact after bulkload, but it used the `selectNow` 
compaction which always select compact files, and set a relatively high 
priority for the compaction. 

We can reduce the compaction pressure after bulk load by avoid repeatedly 
adding same store compactions to the queue after bulk load files.

 


> Avoid repeatedly trigger same store compactions after bulk load files
> ---------------------------------------------------------------------
>
>                 Key: HBASE-26249
>                 URL: https://issues.apache.org/jira/browse/HBASE-26249
>             Project: HBase
>          Issue Type: Improvement
>            Reporter: Xiaolin Ha
>            Assignee: Xiaolin Ha
>            Priority: Major
>
> HBASE-25213 introduced to trigger compact after bulkload, but if a store 
> needs compaction, each file loaded to this store will add a compaction runner 
> of this store to the compaction queue. Then repeatedly compute for 
> needCompaction and selectFiles for this store, and the compaction of the 
> store may be continuous.
> We can mark the queued compaction runners for the stores, and avoid 
> repeatedly adding same store compactions to the queue after bulk load files. 
> This method can reduce the pressure during the period just after bulk loading 
> and spread the pressure to other times by the periodic CompactionChecker.
>  



--
This message was sent by Atlassian Jira
(v8.20.1#820001)

Reply via email to