[
https://issues.apache.org/jira/browse/MESOS-10049?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17099630#comment-17099630
]
Qian Zhang commented on MESOS-10049:
------------------------------------
All the code changes about the newly introduced task status reason
`REASON_CONTAINER_MEMORY_REQUEST_EXCEEDED` (i.e., the above two patches) have
been reverted, so the patch below for details.
commit 6bb60a4869394f663a09370016127ae8688cbe06
Author: Qian Zhang
Date: Mon Apr 27 22:34:51 2020 +0800
Reverted the changes about `REASON_CONTAINER_MEMORY_REQUEST_EXCEEDED`.
The method `MemorySubsystemProcess::oomWaited()` will only be invoked when the
container is OOM killed because it uses more memory than its hard memory limit
(i.e., the task status reason `REASON_CONTAINER_LIMITATION_MEMORY`), it will
NOT be invoked when a burstable container is OOM killed because the agent host
is running out of memory, i.e., we will NOT receive OOM killing notification
via cgroups notification API for this case. So it is not possible for Mesos to
provide a task status reason `REASON_CONTAINER_MEMORY_REQUEST_EXCEEDED` for
this
case.
Review: https://reviews.apache.org/r/72442
> Add a new reason in `TaskStatus::Reason` for the case that a task is
> OOM-killed due to exceeding its memory request
> -------------------------------------------------------------------------------------------------------------------
>
> Key: MESOS-10049
> URL: https://issues.apache.org/jira/browse/MESOS-10049
> Project: Mesos
> Issue Type: Task
> Reporter: Qian Zhang
> Assignee: Greg Mann
> Priority: Major
> Fix For: 1.10.0
>
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
