cloudstack-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <>
Subject [jira] [Commented] (CLOUDSTACK-9407) vm_network_map table doesnt get cleaned up properly
Date Tue, 21 Jun 2016 16:54:57 GMT


ASF GitHub Bot commented on CLOUDSTACK-9407:

Github user nvazquez commented on a diff in the pull request:
    --- Diff: engine/orchestration/src/com/cloud/vm/ ---
    @@ -1577,8 +1577,9 @@ private void advanceStop(final VMInstanceVO vm, final boolean cleanUpEvenIfUnabl
             try {
    -            _networkMgr.release(profile, cleanUpEvenIfUnableToStop);
    -            s_logger.debug("Successfully released network resources for the vm " + vm);
    +            s_logger.debug("Not releasing network resources until expunge command is
    --- End diff --
    I decided to remove try-block as it only logged info, and include it on method's javadoc,
as I don't think it could be worth logging. Do you agree?

> vm_network_map table doesnt get cleaned up properly
> ---------------------------------------------------
>                 Key: CLOUDSTACK-9407
>                 URL:
>             Project: CloudStack
>          Issue Type: Bug
>      Security Level: Public(Anyone can view this level - this is the default.) 
>    Affects Versions: 4.9.0
>            Reporter: Nicolas Vazquez
>            Assignee: Nicolas Vazquez
>             Fix For: 4.9.0
> h3. Introduction
> It was found out that in production environments {{vm_network_map}} table entries were
slowly growing. It was investigated how this entries were cleaned up.
> h3. Behaviour
> On vm creation, vm mappings are inserted on {{vm_network_map}}.
> On vm stop, mappings are deleted from {{vm_network_map}} for vm, as a result of the release
of its nics.
> h3. Problem
> If created vm is stopped from hypervisor side (at least on vSphere in which we tested
it), when CloudStack realizes vm is stopped it doesn't clean up {{vm_network_table}}, and,
as cleanup is made during vm stop, when vm is eventually destroyed and expunged it won't clean
up their entries in that table.
> h3. Proposed solution
> We propose to move {{vm_network_map}} table cleanup to expunge command instead of stop

This message was sent by Atlassian JIRA

View raw message