[
https://issues.apache.org/jira/browse/MESOS-9950?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17369759#comment-17369759
]
Subhajit Palit commented on MESOS-9950:
---------------------------------------
Thanks for that option [~cf.natali] - I will try it out further and share my
observation.
> memory cgroup gone before isolator cleaning up
> ----------------------------------------------
>
> Key: MESOS-9950
> URL: https://issues.apache.org/jira/browse/MESOS-9950
> Project: Mesos
> Issue Type: Bug
> Components: containerization
> Reporter: longfei
> Priority: Major
>
> The memcg created by mesos may have been deleted before cgroup/memory
> isolator cleaning up.
> This would let the termination fail and lose information in the old
> termination(before fail).
> {code:java}
> I0821 15:16:03.025796 3354800 paths.cpp:745] Creating sandbox
> '/opt/tiger/mesos_deploy_videoarch/mesos_zeus/slave/slaves/fb5c1a5b-e106-47c1-9fe3-6ebd311b30ee-S628/frameworks/8e4967e5-736e-4a22-90c3-7b32d526914d-0000/executors/mt:z00000000000003584687:1/runs/a0706ca0-fe2c-4477-8161-329b26ea5d89'
> for user 'tiger'
> I0821 15:16:03.026199 3354800 paths.cpp:748] Creating sandbox
> '/opt/tiger/mesos_deploy_videoarch/mesos_zeus/slave/meta/slaves/fb5c1a5b-e106-47c1-9fe3-6ebd311b30ee-S628/frameworks/8e4967e5-736e-4a22-90c3-7b32d526914d-0000/executors/mt:z00000000000003584687:1/runs/a0706ca0-fe2c-4477-8161-329b26ea5d89'
> I0821 15:16:03.026304 3354800 slave.cpp:9064] Launching executor
> 'mt:z00000000000003584687:1' of framework
> 8e4967e5-736e-4a22-90c3-7b32d526914d-0000 with resources
> [{"allocation_info":{"role":"*"},"name":"cpus","scalar":{"value":0.1},"type":"SCALAR"},{"allocation_info":{"role":"*"},"name":"mem","scalar":{"value":32.0},"type":"SCALAR"}]
> in work directory
> '/opt/tiger/mesos_deploy_videoarch/mesos_zeus/slave/slaves/fb5c1a5b-e106-47c1-9fe3-6ebd311b30ee-S628/frameworks/8e4967e5-736e-4a22-90c3-7b32d526914d-0000/executors/mt:z00000000000003584687:1/runs/a0706ca0-fe2c-4477-8161-329b26ea5d89'
> I0821 15:16:03.051795 3354800 slave.cpp:3520] Launching container
> a0706ca0-fe2c-4477-8161-329b26ea5d89 for executor
> 'mt:z00000000000003584687:1' of framework
> 8e4967e5-736e-4a22-90c3-7b32d526914d-0000
> I0821 15:16:03.076608 3354807 containerizer.cpp:1325] Starting container
> a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:16:03.076911 3354807 containerizer.cpp:3185] Transitioning the state
> of container a0706ca0-fe2c-4477-8161-329b26ea5d89 from PROVISIONING to
> PREPARING
> I0821 15:16:03.077906 3354802 memory.cpp:478] Started listening for OOM
> events for container a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:16:03.079540 3354804 memory.cpp:198] Updated
> 'memory.soft_limit_in_bytes' to 4032MB for container
> a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:16:03.079587 3354820 cpu.cpp:92] Updated 'cpu.shares' to 1126 (cpus
> 1.1) for container a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:16:03.079589 3354804 memory.cpp:227] Updated 'memory.limit_in_bytes'
> to 4032MB for container a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:16:03.080901 3354802 switchboard.cpp:316] Container logger module
> finished preparing container a0706ca0-fe2c-4477-8161-329b26ea5d89;
> IOSwitchboard server is not required
> I0821 15:16:03.081593 3354801 linux_launcher.cpp:492] Launching container
> a0706ca0-fe2c-4477-8161-329b26ea5d89 and cloning with namespaces
> I0821 15:16:03.083823 3354808 containerizer.cpp:2107] Checkpointing
> container's forked pid 1857418 to
> '/opt/tiger/mesos_deploy_videoarch/mesos_zeus/slave/meta/slaves/fb5c1a5b-e106-47c1-9fe3-6ebd311b30ee-S628/frameworks/8e4967e5-736e-4a22-90c3-7b32d526914d-0000/executors/mt:z00000000000003584687:1/runs/a0706ca0-fe2c-4477-8161-329b26ea5d89/pids/forked.pid'
> I0821 15:16:03.084156 3354808 containerizer.cpp:3185] Transitioning the state
> of container a0706ca0-fe2c-4477-8161-329b26ea5d89 from PREPARING to ISOLATING
> I0821 15:16:03.091468 3354808 containerizer.cpp:3185] Transitioning the state
> of container a0706ca0-fe2c-4477-8161-329b26ea5d89 from ISOLATING to FETCHING
> I0821 15:16:03.094933 3354808 containerizer.cpp:3185] Transitioning the state
> of container a0706ca0-fe2c-4477-8161-329b26ea5d89 from FETCHING to RUNNING
> I0821 15:16:03.197753 3354808 memory.cpp:198] Updated
> 'memory.soft_limit_in_bytes' to 4032MB for container
> a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:16:03.197757 3354801 cpu.cpp:92] Updated 'cpu.shares' to 1126 (cpus
> 1.1) for container a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:21:39.692978 3354814 memory.cpp:515] OOM detected for container
> a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:21:39.693182 3354805 containerizer.cpp:3044] Container
> a0706ca0-fe2c-4477-8161-329b26ea5d89 has reached its limit for resource []
> and will be terminated
> I0821 15:21:39.693192 3354805 containerizer.cpp:2518] Destroying container
> a0706ca0-fe2c-4477-8161-329b26ea5d89 in RUNNING state
> I0821 15:21:39.693197 3354805 containerizer.cpp:3185] Transitioning the state
> of container a0706ca0-fe2c-4477-8161-329b26ea5d89 from RUNNING to DESTROYING
> I0821 15:21:39.693542 3354815 linux_launcher.cpp:576] Asked to destroy
> container a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:21:39.693563 3354815 linux_launcher.cpp:618] Destroying cgroup
> '/sys/fs/cgroup/freezer/tiger/a0706ca0-fe2c-4477-8161-329b26ea5d89'
> I0821 15:21:39.693737 3354825 cgroups.cpp:2854] Freezing cgroup
> /sys/fs/cgroup/freezer/tiger/a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:21:39.794579 3354827 cgroups.cpp:1242] Successfully froze cgroup
> /sys/fs/cgroup/freezer/tiger/a0706ca0-fe2c-4477-8161-329b26ea5d89 after
> 100.765952ms
> I0821 15:21:39.795117 3354800 cgroups.cpp:2872] Thawing cgroup
> /sys/fs/cgroup/freezer/tiger/a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:21:39.795686 3354800 cgroups.cpp:1271] Successfully thawed cgroup
> /sys/fs/cgroup/freezer/tiger/a0706ca0-fe2c-4477-8161-329b26ea5d89 after
> 515072ns
> I0821 15:21:39.847966 3354798 containerizer.cpp:3024] Container
> a0706ca0-fe2c-4477-8161-329b26ea5d89 has exited
> E0821 15:21:39.950544 3354823 slave.cpp:6302] Termination of executor
> 'mt:z00000000000003584687:1' of framework
> 8e4967e5-736e-4a22-90c3-7b32d526914d-0000 failed: Failed to clean up an
> isolator when destroying container: Failed to destroy cgroups: Failed to get
> nested cgroups: Failed to determine canonical path of
> '/sys/fs/cgroup/memory/tiger/a0706ca0-fe2c-4477-8161-329b26ea5d89': No such
> file or directory
> W0821 15:21:39.950760 3354817 containerizer.cpp:2462] Skipping status for
> container a0706ca0-fe2c-4477-8161-329b26ea5d89 because: Container does not
> exist
> W0821 15:21:39.950958 3354811 containerizer.cpp:2324] Ignoring update for
> currently being destroyed container a0706ca0-fe2c-4477-8161-329b26ea5d89
> I0821 15:21:39.953343 3354816 gc.cpp:95] Scheduling
> '/opt/tiger/mesos_deploy_videoarch/mesos_zeus/slave/meta/slaves/fb5c1a5b-e106-47c1-9fe3-6ebd311b30ee-S628/frameworks/8e4967e5-736e-4a22-90c3-7b32d526914d-0000/executors/mt:z00000000000003584687:1/runs/a0706ca0-fe2c-4477-8161-329b26ea5d89/tasks/mt:z00000000000003584687:1'
> for gc 6.99610007729185days in the future
> I0821 15:21:39.953478 3354806 gc.cpp:95] Scheduling
> '/opt/tiger/mesos_deploy_videoarch/mesos_zeus/slave/slaves/fb5c1a5b-e106-47c1-9fe3-6ebd311b30ee-S628/frameworks/8e4967e5-736e-4a22-90c3-7b32d526914d-0000/executors/mt:z00000000000003584687:1/runs/a0706ca0-fe2c-4477-8161-329b26ea5d89'
> for gc 6.99998896460444days in the future
> I0821 15:21:39.953538 3354810 gc.cpp:95] Scheduling
> '/opt/tiger/mesos_deploy_videoarch/mesos_zeus/slave/meta/slaves/fb5c1a5b-e106-47c1-9fe3-6ebd311b30ee-S628/frameworks/8e4967e5-736e-4a22-90c3-7b32d526914d-0000/executors/mt:z00000000000003584687:1/runs/a0706ca0-fe2c-4477-8161-329b26ea5d89'
> for gc 6.99998896395852days in the future
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)