You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@dolphinscheduler.apache.org by "lgcareer2019@outlook.com" <lg...@outlook.com> on 2020/04/19 02:42:15 UTC

回复: Re: [DISCUSS] fault-tolerant on dev

It's great!



DolphinScheduler(Incubator) PPMC
Gang Li 李岗

lgcareer2019@outlook.com<ma...@outlook.com>
 
发件人: qiaozhanwei@outlook.com
发送时间: 2020-04-16 11:41
收件人: dev
主题: Re: [DISCUSS] fault-tolerant on dev
https://github.com/apache/incubator-dolphinscheduler/pull/2420 
 
this PR sloved sync wait when no master or worker
 
―――――――――――――
DolphinScheduler(Incubator)  PPMC
Zhanwei Qiao 乔占卫
 
qiaozhanwei@outlook.com
From: leon bao
Date: 2020-04-13 16:56
To: dev
Subject: [DISCUSS] fault-tolerant on dev
At present, the master and worker on the dev branch rely on each other's
communication to run tasks and modify task status. There are also two
problems:
1. The normal task execution of the worker ends, but the master does not
exist, or the master fails after receiving the information reported by the
worker, at present, the information from worker will be lost.
2. When the master runs normally, but the worker does not exist, how to
deal with the task list to be run?
-- 
DolphinScheduler(Incubator)  PPMC
BaoLiang 鲍亮
leonbao@apache.org