You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@zeppelin.apache.org by Ruslan Dautkhanov <da...@gmail.com> on 2018/02/12 23:45:31 UTC
TimeoutLifecycleManager
Trying out below settings for the new TimeoutLifecycleManager :
> zeppelin.interpreter.lifecyclemanager.class
> org.apache.zeppelin.interpreter.lifecycle.TimeoutLifecycleManager
> zeppelin.interpreter.lifecyclemanager.timeout.checkinterval
> 60000
> zeppelin.interpreter.lifecyclemanager.timeout.threshold
> 600000
>
My understanding is should have killed my Spark interpreter in 10 minutes
of idleness (given zeppelin.interpreter.lifecyclemanager.timeout.threshold
= 600000 ms above)
but for some reason it keeps running.. am I missing something?
No messages of attempts to timeout the interpreter in the logs even at
DEBUG level.
Thanks,
Ruslan Dautkhanov
Re: TimeoutLifecycleManager
Posted by Ruslan Dautkhanov <da...@gmail.com>.
Nevermind. I figured this out. It works as expected. Great feature!
--
Ruslan Dautkhanov
On Mon, Feb 12, 2018 at 5:16 PM, Jeff Zhang <zj...@gmail.com> wrote:
>
> Right, you spark interpreter should be killed in 10 minutes of idless,
> could you attach your full zeppelin server log & spark interpreter log ?
>
>
> Ruslan Dautkhanov <da...@gmail.com>于2018年2月13日周二 上午7:45写道：
>
>> Trying out below settings for the new TimeoutLifecycleManager :
>>
>>
>>> zeppelin.interpreter.lifecyclemanager.class
>>> org.apache.zeppelin.interpreter.lifecycle.TimeoutLifecycleManager
>>> zeppelin.interpreter.lifecyclemanager.timeout.checkinterval
>>> 60000
>>> zeppelin.interpreter.lifecyclemanager.timeout.threshold
>>> 600000
>>>
>>
>> My understanding is should have killed my Spark interpreter in 10 minutes
>> of idleness (given zeppelin.interpreter.lifecyclemanager.timeout.
>> threshold = 600000 ms above)
>> but for some reason it keeps running.. am I missing something?
>>
>> No messages of attempts to timeout the interpreter in the logs even at
>> DEBUG level.
>>
>>
>> Thanks,
>> Ruslan Dautkhanov
>>
>
Re: TimeoutLifecycleManager
Posted by Jeff Zhang <zj...@gmail.com>.
Right, you spark interpreter should be killed in 10 minutes of idless,
could you attach your full zeppelin server log & spark interpreter log ?
Ruslan Dautkhanov <da...@gmail.com>于2018年2月13日周二 上午7:45写道：
> Trying out below settings for the new TimeoutLifecycleManager :
>
>
>> zeppelin.interpreter.lifecyclemanager.class
>> org.apache.zeppelin.interpreter.lifecycle.TimeoutLifecycleManager
>> zeppelin.interpreter.lifecyclemanager.timeout.checkinterval
>> 60000
>> zeppelin.interpreter.lifecyclemanager.timeout.threshold
>> 600000
>>
>
> My understanding is should have killed my Spark interpreter in 10 minutes
> of idleness (given
> zeppelin.interpreter.lifecyclemanager.timeout.threshold = 600000 ms
> above)
> but for some reason it keeps running.. am I missing something?
>
> No messages of attempts to timeout the interpreter in the logs even at
> DEBUG level.
>
>
> Thanks,
> Ruslan Dautkhanov
>