[jira] [Comment Edited] (MESOS-5879) cgroups/net_cls isolator causing agent recovery issues

["Avinash Sridharan (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/MESOS-5879?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15388159#comment-15388159 ] 

Avinash Sridharan edited comment on MESOS-5879 at 7/21/16 6:36 PM:
-------------------------------------------------------------------

Are you seeing the following error:

```
        "The secondary handle " + hexify(handle.secondary) +
        ", for the primary handle " + hexify(handle.primary) +
        " has already been allocated");
```

This should never happen, unless your custom isolator was going and modifying the net_cls handles as well ? If no other actor is modifying these handles except the `NetClsHandleManager` then we shouldn't see this problem. Need to see if its some other bug that might be manifesting here. Given that we should never see this problem if there is a single entity allocating the net_cls handles, don't thing blindly destroying the orphan container might not be the right solution here.


was (Author: avinash@mesosphere.io):
Are you seeing the following error:

```
        "The secondary handle " + hexify(handle.secondary) +
        ", for the primary handle " + hexify(handle.primary) +
        " has already been allocated");
```

This should never happen, unless your custom isolator was going and modifying the net_cls handles as well ? If no other actor is modifying these handles except the `NetClsHandleManager` then we shouldn't see this problem, and there might be a different bug here, then just blindly requiring to destroy the orphan container.

> cgroups/net_cls isolator causing agent recovery issues
> ------------------------------------------------------
>
>                 Key: MESOS-5879
>                 URL: https://issues.apache.org/jira/browse/MESOS-5879
>             Project: Mesos
>          Issue Type: Bug
>          Components: cgroups, isolation, slave
>            Reporter: Silas Snider
>            Assignee: Avinash Sridharan
>
> We run with 'cgroups/net_cls' in our isolator list, and when we restart any agent process in a cluster running an experimental custom isolator as well, the agents are unable to recover from checkpoint, because net_cls reports that unknown orphan containers have duplicate net_cls handles.
> While this is a problem that needs to be solved (probably by fixing our custom isolator), it's also a problem that the net_cls isolator fails recovery just for duplicate handles in cgroups that it is literally about to unconditionally destroy during recovery. Can this be fixed?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)