[
https://issues.apache.org/jira/browse/HBASE-27216?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17678976#comment-17678976
]
Duo Zhang commented on HBASE-27216:
-----------------------------------
Oh, just cleaning the hbase:replication table is not enough, in SCP, before
claiming replication queues for a dead server, we will try to add the missing
replication queues. The logic is that, every peer should have a replication
queue, so if we delete all data in hase:replication, SCP will add a replication
queue for every replication peer, and then we will replicate all the data
again...
Let me think how to fix this... WIll post there later when I find a way
[~heliangjun] [~liyu]
> Revisit the ReplicationSyncUp tool
> ----------------------------------
>
> Key: HBASE-27216
> URL: https://issues.apache.org/jira/browse/HBASE-27216
> Project: HBase
> Issue Type: Sub-task
> Components: Replication, tooling
> Reporter: Duo Zhang
> Assignee: Liangjun He
> Priority: Major
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
