[
https://issues.apache.org/jira/browse/CLOUDSTACK-9650?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15817529#comment-15817529
]
ASF GitHub Bot commented on CLOUDSTACK-9650:
--------------------------------------------
Github user rhtyd commented on a diff in the pull request:
https://github.com/apache/cloudstack/pull/1899#discussion_r95524526
--- Diff: api/src/com/cloud/deploy/DeploymentClusterPlanner.java ---
@@ -55,7 +55,7 @@
"true",
"Enable/Disable cluster thresholds. If disabled, an instance
can start in a cluster even though the threshold may be crossed.",
false,
- ConfigKey.Scope.Zone);
+ ConfigKey.Scope.Global);
--- End diff --
@koushik-das Ignore, the above comment. But, I think making this global
takes away the option to override this at a zone level? Why remove this?
> Allow starting VMs regardless of cpu/memory cluster.disablethreshold setting
> ----------------------------------------------------------------------------
>
> Key: CLOUDSTACK-9650
> URL: https://issues.apache.org/jira/browse/CLOUDSTACK-9650
> Project: CloudStack
> Issue Type: Improvement
> Security Level: Public(Anyone can view this level - this is the
> default.)
> Components: Management Server
> Affects Versions: 4.10.0.0
> Reporter: Koushik Das
> Assignee: Koushik Das
> Fix For: 4.10.1.0
>
>
> VM deployments are not allowed on clusters where resource (cpu/memory)
> allocation has exceeded the cluster disabled thresholds. The same policy also
> gets applied in case of start VM if last host where the VM was running
> doesn't have enough capacity and a new host is picked up. In certain
> scenarios this can be restrictive and despite having capacity, the VM cannot
> be started.
> This improvement is to provide administrator an option to disable/enable
> cluster threshold enforcement during start of a stopped VM as long as
> sufficient capacity is available.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
