[
https://issues.apache.org/jira/browse/HBASE-24265?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17093172#comment-17093172
]
Duo Zhang edited comment on HBASE-24265 at 4/27/20, 8:26 AM:
-------------------------------------------------------------
Oh no, we have this in code...
{code}
CompletableFuture.runAsync(
() -> doCallMethod(method, (HBaseRpcController)controller, request,
responsePrototype, done));
{code}
It will execute the method in ForkJoinPool.commonPool, which will be shared by
lots of tasks, and easy to lead to a dead lock... And this is not the correct
way to implement an asynchronous operation...
was (Author: apache9):
Oh no, we have this in code...
{noformat}
CompletableFuture.runAsync(
() -> doCallMethod(method, (HBaseRpcController)controller, request,
responsePrototype, done));
{noformat}
It will execute the method in ForkJoinPool.commonPool, which will be shared by
lots of tasks, and easy to lead to a dead lock... And this is not the correct
way to implement an asynchronous operation...
> Remove hedged rpc call support, implement the logic in MaterRegistry directly
> -----------------------------------------------------------------------------
>
> Key: HBASE-24265
> URL: https://issues.apache.org/jira/browse/HBASE-24265
> Project: HBase
> Issue Type: Sub-task
> Components: IPC/RPC
> Reporter: Duo Zhang
> Assignee: Duo Zhang
> Priority: Major
> Fix For: 3.0.0
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
