[jira] [Commented] (HBASE-27189) NettyServerRpcConnection is properly closed when the netty channel is closed

["Duo Zhang (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HBASE-27189?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17564696#comment-17564696 ] 

Duo Zhang commented on HBASE-27189:
-----------------------------------

[~apurtell] [~vjasani] FYI.

Found this when trying to implement HBASE-27185.

I've already implement a UT to reproduce the problem, see this PR https://github.com/apache/hbase/pull/4611.

Let me think whether we need to do some refactoring before fixing it, to make the code cleaner, or just a simple change to fix the problem first...

> NettyServerRpcConnection is properly closed when the netty channel is closed
> ----------------------------------------------------------------------------
>
>                 Key: HBASE-27189
>                 URL: https://issues.apache.org/jira/browse/HBASE-27189
>             Project: HBase
>          Issue Type: Bug
>          Components: netty, rpc
>            Reporter: Duo Zhang
>            Assignee: Duo Zhang
>            Priority: Blocker
>             Fix For: 2.5.0, 3.0.0-alpha-4, 2.4.14
>
>
> We have a close method for NettyServerRpcConnection where we dispose the SaslServer and also release the ByteBuf if needed, but we never call it...
> This could also leak memory and also other resources...



--
This message was sent by Atlassian Jira
(v8.20.10#820010)