2012年6月4日 下午 12:42
our hyper-v cluster was previously managed via scvmm 2008 r2. This has been de-installed (including SQL-DB) and new SCVMM 2012 RTM installed. The "old" service account has been recycled. The account is a domain user and a member of "Local Admin Group" on SCVMM server and all Host servers. The cluster was added.
Error one: in SCVMM I always get the message "cluster overcommitted". under property the values on the host folder is default and has not been changed for hosts. ALL VMs are offline. Still the message cluster overcommited is shown. I have no idea why SCVMM gives me this message when no VM is running. Under "fabric" the host servers are shown with correct values (free memory...). I looked in Event Viewer to see if there are any error messages showing an error... I have not found anything.
Error two: I cannot deploy a VM. I tried creating from template, from vhd, from scratch... I am not able to place the VM on a Host. I always receive the "over commit" error and error message "configuration issues related to the virtual hard drive configuration with ID [ID Number] of virtual machine [name] prevent deployment". Now I tried to create a VM as HA with CSV as end path and choosing a host from the cluster. I tried to create a VM non-HA with local storage as path... Never works, always same error message.
Strange thing is: creating/moving a HA VM from Hyper-V Manager works like a charm - no problems at all. Only when trying to use SCVMM 2012 do I see those errors.
Any ideas what I could do to troubleshoot the issue?
2012年6月4日 下午 05:03
Cluster overcommitment is a VMM-level calculation which attempts to answer the question, "If X nodes simultaneously fail, will failover cluster manager always be able to restart all VMs on working nodes". It's a complex calculation that you can read more about at this VMM blog.
Stopped VMs are considered to use resources for the cluster overcommit calculation. The reason is that we assume that any HA VM deployed to a cluster could be started by an SSU or out of band, and cluster overcommit has to consider the worst case scenario. If you want to store a VM, consider migrating it to a library so it doesn't consume resources on the cluster itself.
How can you work around this? Either store some VMs to the library, or you can simply switch off the cluster overcommit check, available under the cluster properties in VMM. (Set cluster reserve to 0).
2012年6月5日 下午 02:19
Thank you for that responce, that resolved one of my questions. however, there was an additional question asked that I also need an answer to.
"Error two: I cannot deploy a VM. I tried creating from template, from vhd, from scratch... I am not able to place the VM on a Host. I always receive the "over commit" error and error message "configuration issues related to the virtual hard drive configuration with ID [ID Number] of virtual machine [name] prevent deployment". Now I tried to create a VM as HA with CSV as end path and choosing a host from the cluster. I tried to create a VM non-HA with local storage as path... Never works, always same error message." by SROE
I only have one server setup with clustering right now, as I'm trying to migrate the company - I recently was hired at - off local storage. I am not able to put anythng on my clustered storage sense I migrated our SCVMM from 2008 R2 to 2012. Here is the error I get when trying to migrate a template from a local storage hyper-v server to a clustered storage hyper-v server.
Cannot create or update a non highly available virtual machine because the path C:\ClusterStorage\Volume2 is a clustered resource.
Ensure that the path exists and that the path is not a cluster disk or cluster shared volume, and then try the operation again.
2012年6月25日 上午 08:59版主
okay. as the error state you can not place non highly available virtual machine on CSV "Since this storage is shared and can disappear from that host any time so the VM will have local resources "Settings" without VHD file"
VMM 2012 gives you the ability to host local Machines on local storage and HA machine in CSV.
Mohamed Fawzi | http://fawzi.wordpress.com
- 已標示為解答 Kristian NeseMVP, Moderator 2012年7月2日 下午 06:19
2012年7月31日 下午 07:29
Has anyone found an answer to this?
I'm having this exact same issue after migrating our clusters from VMM 2008 R2 to VMM 2012. It only impacts some of our clusters, but I have yet to find a difference between the clusters. I cannot deploy or clone new VMs to our CSVs - I also receive the "configuration issues related to the virtual hard drive configuration with ID [ID Number] of virtual machine [name] prevent deployment" message. I too am able to deploy new VMs in Hyper-V manager without issue, but they are not HA until migrated in VMM to make them HA.
2012年8月21日 下午 07:07Has anyone ever gotten this resolved? We have the same issue.
2012年8月21日 下午 11:56
The underlying issue in this original question was cluster overcommit. The default cluster reserve is 1. Setting this to 0 will allow new VMs to be created. If this doesn't resolve your issue, please provide further details of exactly what error you are seeing.
2013年2月25日 下午 09:05
OK so I have this same issue and I keep reading the exact same response but it does not help us fix the issue. I have VMs that are currently being hosted in a non-HA environment. I am trying to move them to an HA environment that I have built out. Is there a way to tell VMM to make the VM HA upon moving it? I don't swee this anywhere and noone seems to discuss it. This was never an issue in VMM 2007 and I had used it very reliably in the past.