You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@dolphinscheduler.apache.org by wenhemin <wh...@163.com> on 2020/04/14 03:22:45 UTC

[DISCUSS] When task retry strategy, merge task instance

hello, friends!

When task retry strategy, it is recommended to merge task instance.
At present, retry every time, a new task instance will be created, and increase retry count by one.
I think, no need to create a new task every time, only increase retry count by one.
The purpose of this is to facilitate query and statistics.

Now:

Re: [DISCUSS] When task retry strategy, merge task instance

Posted by wenhemin <wh...@163.com>.
If add condition by retry_times = 0, I won't see any subsequent retry tasks,
Can't see the current status of the retry task.

I think retrying the task should not create a new task.

> 在 2020年4月14日,14:07,lidong dai <da...@gmail.com> 写道:
> 
> good idea,  but I think task instance should generate a new one every time when it indeed runs. If there is no newly generated task instance, some person maybe will think there is a problem with the system. for your situation maybe you could  add the where condition by retry_times = 0?
>     
> 
> 
> 
> Best Regards
> ---------------
> DolphinScheduler(Incubator) PPMC
> Lidong Dai 代立冬
> dailidong66@gmail.com <ma...@gmail.com>
> ---------------
> 
> 
> wenhemin <whm_777@163.com <ma...@163.com>> 于2020年4月14日周二 上午11:23写道:
> hello, friends!
> 
> When task retry strategy, it is recommended to merge task instance.
> At present, retry every time, a new task instance will be created, and increase retry count by one.
> I think, no need to create a new task every time, only increase retry count by one.
> The purpose of this is to facilitate query and statistics.
> 
> Now:
> 


Re: [DISCUSS] When task retry strategy, merge task instance

Posted by lidong dai <da...@gmail.com>.
good idea,  but I think task instance should generate a new one every time
when it indeed runs. If there is no newly generated task instance, some
person maybe will think there is a problem with the system. for your
situation maybe you could  add the where condition by retry_times = 0?




Best Regards
---------------
DolphinScheduler(Incubator) PPMC
Lidong Dai 代立冬
dailidong66@gmail.com
---------------


wenhemin <wh...@163.com> 于2020年4月14日周二 上午11:23写道:

> hello, friends!
>
> When task retry strategy, it is recommended to merge task instance.
> At present, retry every time, a new task instance will be created, and
> increase retry count by one.
> I think, no need to create a new task every time, only increase retry
> count by one.
> The purpose of this is to facilitate query and statistics.
>
> Now:
>