[
https://issues.apache.org/jira/browse/HBASE-18624?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16133601#comment-16133601
]
Ajay Jadhav commented on HBASE-18624:
-------------------------------------
[~anoop.hbase]: I think this makes sense even for non-replica clusters. It is a
simple utility API to clear the blockcache based on tablename. We found this
useful even in our performance testing instead of restarting the HBase cluster
every time. Let me know if you have any concerns.
[~ram_krish]: In the case of replica(s), all the clusters are pointing to same
rootDir. So, whenever compaction runs, the archived files will be cleared by
cleaner chore on the primary cluster itself.
Having said that, we did provide a mechanism to disable the cleaner during
HBase on S3 work to avoid performance overhead during peak times.
[~gsbiju]: Even though this is part of the work done for the parent JIRA, it
can be useful in other cases too. The umbrella Jira is where we are proposing
the new design for read replica clusters i.e. primary HBase cluster which
allows write/ read, and replica(s) (which are separate HBase clusters),
allowing only reads. The only thing that is common is that they share a common
root dir/ S3 bucket for data.
> Added support for clearing BlockCache
> -------------------------------------
>
> Key: HBASE-18624
> URL: https://issues.apache.org/jira/browse/HBASE-18624
> Project: HBase
> Issue Type: Sub-task
> Affects Versions: 2.0.0, 1.3.0
> Reporter: Ajay Jadhav
> Assignee: Ajay Jadhav
> Priority: Minor
> Fix For: 1.4.0
>
>
> Bulk loading the primary HBase cluster triggers a lot of compactions
> resulting in archival/ creation
> of multiple HFiles. This process will cause a lot of items to become stale in
> replica’s BlockCache.
> This patch will help users to clear the block cache for a given table by
> either using shell or API.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
