[
https://issues.apache.org/jira/browse/CLOUDSTACK-4881?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13862662#comment-13862662
]
Nitin Mehta commented on CLOUDSTACK-4881:
-----------------------------------------
Marcus - To your question - "Who decided it was right to only allow editing the
offering while the VM was off?" -
Let me just tell you what the behavior is.
changeServiceForVirtualMachine - should change the service offering of a
stopped vm and has been in CS since 3.0.x or before. Recently there was a bug
introduced where it was allowing to change the service offering of a running vm
as well. That was never intended and breaks compatibility and tests and has
been fixed with this bug. This api will be deprecated in 5.0
scaleVirtualMachine - introduced in 4.2, this should allow changing the service
offering of a running as well as stopped vm. You can use this api to edit the
offering whenever you want.
If you were using changeServiceForVirtualMachine for changing the offering of a
running and stopped vm, I would say please change that and start using
scaleVirtualMachine instead because changeServiceForVirtualMachine has been
since ages only supporting editing the offering for a stopped vm only.
> Both Running and stopped vm can be scaled up using old API
> changeServiceForVirtualMachine that too above host capacity
> ----------------------------------------------------------------------------------------------------------------------
>
> Key: CLOUDSTACK-4881
> URL: https://issues.apache.org/jira/browse/CLOUDSTACK-4881
> Project: CloudStack
> Issue Type: Bug
> Security Level: Public(Anyone can view this level - this is the
> default.)
> Components: Automation
> Affects Versions: 4.3.0
> Reporter: Gaurav Aradhye
> Assignee: Nitin Mehta
> Priority: Blocker
> Fix For: 4.3.0
>
>
> changeServiceForVirtualMachine API can be used to scale up a running vm also
> (It should be effective against only stopped vm)
> More over, using this API, VM can be scaled up above host capacity (both CPU
> and RAM) without any over-provisioning of resources.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
