You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@cloudstack.apache.org by "Prieur Leary (JIRA)" <ji...@apache.org> on 2014/04/04 07:17:14 UTC

[jira] [Comment Edited] (CLOUDSTACK-6334) Issue is Reproduceable - Suggest Re-open

    [ https://issues.apache.org/jira/browse/CLOUDSTACK-6334?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13959644#comment-13959644 ] 

Prieur Leary edited comment on CLOUDSTACK-6334 at 4/4/14 5:15 AM:
------------------------------------------------------------------

Spent a bit more time on this:

1. I verified the agent is 4.3.0-1.el6  (from Yum list). Management Server is too.
2. I verified that 'vm.memballoon.disable=true' is in the agent.parameters file.
3. I restarted VMs on more than one host, many times. It always shows <memballoon model='none' ><alias name='balloon0'/>.

I was however able to set the memory over provision factor to 2 (in the Cluster settings), coupled with creating a custom system offering with 512MB RAM for the Virtual Router; this allows the router to restart without a problem.

Over provisioning however is an important feature and would like to get to the bottom of this. I did notice that issue 5968 references that the code update for this issue was for 4.4.0. Is it possible the code I am running just does not include this feature?




was (Author: prieur3):
Spent a bit more time on this:

1. I verified the agent is 4.3.0-1.el6  (from Yum list). Management Server is too.
2. I verified that 'vm.memballoon.disable=true' is in the agent.parameters file.
3. I restarted VMs on more than one host, many times. It always shows <memballoon model='none' ><alias name='balloon0'/>.

I was however able to set the memory over provision factor to 2 (in the Cluster settings), coupled with creating a custom system offering with 512MB RAM for the Virtual Router; this allows the router to restart without a problem.

Over provisioning however is an important feature and would like to get to the bottom of this. I did notice that issue 5968 references that the code update for this issue was for 4.4.0. Is it possible the code I am running just does include this feature?



> Issue is Reproduceable - Suggest Re-open 
> -----------------------------------------
>
>                 Key: CLOUDSTACK-6334
>                 URL: https://issues.apache.org/jira/browse/CLOUDSTACK-6334
>             Project: CloudStack
>          Issue Type: Sub-task
>      Security Level: Public(Anyone can view this level - this is the default.) 
>          Components: Virtual Router
>    Affects Versions: 4.3.0
>         Environment: CentOS - KVM
>            Reporter: Prieur Leary
>             Fix For: 4.2.0
>
>
> This is issue is reproducable when setting the cluster property mem.overprovisioning.factor greater than 1.
> If the virtual router is started or re-started and the variable is set greater than one, the Virtual Router VM is created on the host, but never boots.



--
This message was sent by Atlassian JIRA
(v6.2#6252)