[
https://issues.apache.org/jira/browse/HBASE-19617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16303146#comment-16303146
]
Duo Zhang commented on HBASE-19617:
-----------------------------------
I plan to do this after HBASE-19599 is resolved.
And there is another problem, we will block on refreshing peer when adding a
dummy peer. Seems we will create the ConnectionImplementation synchronously and
since the peer dummy, which means the remote does not exist, we will block on
getting cluster id until the retry limit is reached. And if we change to use
AsyncConnection later, the refresh may fail since we will fail the construction
when we can not get the cluster id.
You can see the problem when running TestReplicationAdmin and
TestAsyncAdminReplicationApi. I think we should address this. You can open a
new issue to address it once you confirm the same problem.
Thanks.
> Remove ReplicationQueues, use ReplicationQueueStorage directly
> --------------------------------------------------------------
>
> Key: HBASE-19617
> URL: https://issues.apache.org/jira/browse/HBASE-19617
> Project: HBase
> Issue Type: Sub-task
> Components: Replication
> Reporter: Duo Zhang
>
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
