You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@mesos.apache.org by 陈强 <ch...@qiyi.com> on 2016/01/28 03:20:40 UTC
Docker container does not be killed in process of upgrading.
Hi all,
I met an issue that docker container doesn't be killed in the process of upgrading.
I upgrade mesos from 0.21.0 to 0.25.0, when upgrading a slave, on which a marathon task running (UI would see this).
I expect that the task should be killed and another same marathon task should be started in other slaves. but the task didn't be killed.
If run `docker ps` in all the slaves, would find that there 4 docker container are running that task in the same cluster, but I only started one.
Does someone meet this issue?
Thanks,
Best regards,
Chen, Qiang
Re: Docker container does not be killed in process of upgrading.
Posted by Benjamin Mahler <bm...@apache.org>.
+tim, jie
On Wed, Jan 27, 2016 at 9:48 PM, 陈强 <ch...@qiyi.com> wrote:
> Thanks haosdent.
>
> The issue is tracked with MESOS-3573
> <https://issues.apache.org/jira/browse/MESOS-3573> now, which is still
> unfixed.
>
>
> On 2016年01月28日 10:43, haosdent wrote:
>
> https://issues.apache.org/jira/browse/MESOS-1884
> https://issues.apache.org/jira/browse/MESOS-1915 also may related to your
> problem.
>
> On Thu, Jan 28, 2016 at 10:42 AM, haosdent <ha...@gmail.com> wrote:
>
>> I remember mesos change docker name format. And you skip from 0.21 to
>> 0.25 may cause this problem. And I not sure whether your problem related to
>> this issue or not. https://issues.apache.org/jira/browse/MESOS-3808
>>
>> On Thu, Jan 28, 2016 at 10:20 AM, 陈强 < <ch...@qiyi.com>
>> chenqiang@qiyi.com> wrote:
>>
>>>
>>> Hi all,
>>>
>>> I met an issue that docker container doesn't be killed in the process of
>>> upgrading.
>>>
>>> I upgrade mesos from 0.21.0 to 0.25.0, when upgrading a slave, on which
>>> a marathon task running (UI would see this).
>>> I expect that the task should be killed and another same marathon task
>>> should be started in other slaves. but the task didn't be killed.
>>> If run `docker ps` in all the slaves, would find that there 4 docker
>>> container are running that task in the same cluster, but I only started one.
>>>
>>> Does someone meet this issue?
>>>
>>> Thanks,
>>>
>>> Best regards,
>>> Chen, Qiang
>>>
>>>
>>
>>
>> --
>> Best Regards,
>> Haosdent Huang
>>
>
>
>
> --
> Best Regards,
> Haosdent Huang
>
>
> --
> 陈强
>
> 技术产品中心 计算云 弹性计算
> 手机：＋86 15900964316
> 分机：8377
>
>
Re: Docker container does not be killed in process of upgrading.
Posted by 陈强 <ch...@qiyi.com>.
Thanks haosdent.
The issue is tracked with MESOS-3573
<https://issues.apache.org/jira/browse/MESOS-3573> now, which is still
unfixed.
On 2016年01月28日 10:43, haosdent wrote:
> https://issues.apache.org/jira/browse/MESOS-1884
> https://issues.apache.org/jira/browse/MESOS-1915 also may related to
> your problem.
>
> On Thu, Jan 28, 2016 at 10:42 AM, haosdent <haosdent@gmail.com
> <ma...@gmail.com>> wrote:
>
> I remember mesos change docker name format. And you skip from 0.21
> to 0.25 may cause this problem. And I not sure whether your
> problem related to this issue or not.
> https://issues.apache.org/jira/browse/MESOS-3808
>
> On Thu, Jan 28, 2016 at 10:20 AM, 陈强 <chenqiang@qiyi.com
> <ma...@qiyi.com>> wrote:
>
>
> Hi all,
>
> I met an issue that docker container doesn't be killed in the
> process of upgrading.
>
> I upgrade mesos from 0.21.0 to 0.25.0, when upgrading a slave,
> on which a marathon task running (UI would see this).
> I expect that the task should be killed and another same
> marathon task should be started in other slaves. but the task
> didn't be killed.
> If run `docker ps` in all the slaves, would find that there 4
> docker container are running that task in the same cluster,
> but I only started one.
>
> Does someone meet this issue?
>
> Thanks,
>
> Best regards,
> Chen, Qiang
>
>
>
>
> --
> Best Regards,
> Haosdent Huang
>
>
>
>
> --
> Best Regards,
> Haosdent Huang
--
陈强
技术产品中心 计算云 弹性计算
手机：＋86 15900964316
分机：8377
Re: Docker container does not be killed in process of upgrading.
Posted by haosdent <ha...@gmail.com>.
https://issues.apache.org/jira/browse/MESOS-1884
https://issues.apache.org/jira/browse/MESOS-1915 also may related to your
problem.
On Thu, Jan 28, 2016 at 10:42 AM, haosdent <ha...@gmail.com> wrote:
> I remember mesos change docker name format. And you skip from 0.21 to 0.25
> may cause this problem. And I not sure whether your problem related to this
> issue or not. https://issues.apache.org/jira/browse/MESOS-3808
>
> On Thu, Jan 28, 2016 at 10:20 AM, 陈强 <ch...@qiyi.com> wrote:
>
>>
>> Hi all,
>>
>> I met an issue that docker container doesn't be killed in the process of
>> upgrading.
>>
>> I upgrade mesos from 0.21.0 to 0.25.0, when upgrading a slave, on which a
>> marathon task running (UI would see this).
>> I expect that the task should be killed and another same marathon task
>> should be started in other slaves. but the task didn't be killed.
>> If run `docker ps` in all the slaves, would find that there 4 docker
>> container are running that task in the same cluster, but I only started one.
>>
>> Does someone meet this issue?
>>
>> Thanks,
>>
>> Best regards,
>> Chen, Qiang
>>
>>
>
>
> --
> Best Regards,
> Haosdent Huang
>
--
Best Regards,
Haosdent Huang
Re: Docker container does not be killed in process of upgrading.
Posted by haosdent <ha...@gmail.com>.
I remember mesos change docker name format. And you skip from 0.21 to 0.25
may cause this problem. And I not sure whether your problem related to this
issue or not. https://issues.apache.org/jira/browse/MESOS-3808
On Thu, Jan 28, 2016 at 10:20 AM, 陈强 <ch...@qiyi.com> wrote:
>
> Hi all,
>
> I met an issue that docker container doesn't be killed in the process of
> upgrading.
>
> I upgrade mesos from 0.21.0 to 0.25.0, when upgrading a slave, on which a
> marathon task running (UI would see this).
> I expect that the task should be killed and another same marathon task
> should be started in other slaves. but the task didn't be killed.
> If run `docker ps` in all the slaves, would find that there 4 docker
> container are running that task in the same cluster, but I only started one.
>
> Does someone meet this issue?
>
> Thanks,
>
> Best regards,
> Chen, Qiang
>
>
--
Best Regards,
Haosdent Huang