[
https://issues.apache.org/jira/browse/RATIS-2329?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Xinyu Tan updated RATIS-2329:
-----------------------------
Summary: NettyRpcProxy should support handling netty.channel exceptions to
prevent replication stuck (was: NettyRpcProxy should support handling
netty.channel exceptions.)
> NettyRpcProxy should support handling netty.channel exceptions to prevent
> replication stuck
> -------------------------------------------------------------------------------------------
>
> Key: RATIS-2329
> URL: https://issues.apache.org/jira/browse/RATIS-2329
> Project: Ratis
> Issue Type: Bug
> Components: Netty
> Affects Versions: 2.5.1
> Reporter: Xianming Lei
> Priority: Major
> Time Spent: 1h 50m
> Remaining Estimate: 0h
>
> In our production environment celeborn cluster, due to network problems, a
> TimeOutIOException will occur for a period of time when the leader appends
> entries to the follower. When the network is restored, the leader will be
> stuck when continuing to append entries to the follower, causing the log to
> be unable to be replicated.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)