[
https://issues.apache.org/jira/browse/HBASE-17274?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Phil Yang updated HBASE-17274:
------------------------------
Description:
If we have many peers or some servers have many recovered queues, the total
read size per second from WALs will be very high.
So we should add a server level throttler to limit the total speed of reading
WALs.
was:
If we have many peers or some servers have many recovered queues, the total
read size per second from WALs will be very high, which will increase the
pressure of GC, even maybe OOM because we will read entries for 64MB to buffer
in default.
So we should add a server level throttler to limit the speed of reading WALs,
and limit the total buffered entries for all ReplicationSources.
> Add a server level throttler for replication sources
> ----------------------------------------------------
>
> Key: HBASE-17274
> URL: https://issues.apache.org/jira/browse/HBASE-17274
> Project: HBase
> Issue Type: Improvement
> Components: Replication
> Reporter: Phil Yang
> Assignee: Phil Yang
>
> If we have many peers or some servers have many recovered queues, the total
> read size per second from WALs will be very high.
> So we should add a server level throttler to limit the total speed of reading
> WALs.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
