Trying to recompose some pools but need to do the ones that are not in use. When I start the recompose process I can see the cloning new machine in Vcenter then after that completes it just halts. Nothing changes in Vcentter and nothing changes in View manager. When i look at teh VM in Vcenter the machine is powered off and never comes back on. I waited well over 30 minutes for it to change but nothing. If yI try to manually power teh VM on it powers on but then gets an error "Provisioning error occurred for Machine view004: Refit operation resync failed" After that it continues to recompose the VM but uses the old Image and not teh new one. Tried this on multiple VM's and two different pools. Anyone have an idea what is going on? Running Horizon View 5.2 on ESXi 5.1 U1. Vcenter and Composer server are separated. Any help is appreciated. Thanks.
Task says Recompose and status says Maintenance Mode. I'm trying another but going to wait even longer than 30 minutes. NOt sure why i am doing that but thought i would give it a shot.
What kind of storage do you have? Do you have a large delta file after you changed the parent image? How is your network configured? Are you using shared storage or local storage for your VDI VMs?
We have an Equalogic PS6500 Storage array and it's running on 10GB, Delta is very small, no shared storage all in the same volume. The VDI desktops are setup to use two volumes. We just have the linked clones on their own Volume.
Trying recomposing the pool with the "Use View Storage Accelerator" unchecked. This can be found in the "Advanced Storage Settings" tab when you edit the pool. I was having a similar issue and this helped. We did not see a performance decrease with it unchecked either.
Equalogic PS6500 is iSCSI SAN. Can you check your Physical Switch configuration? How about you use local datastore and deploy the VDI VMs and see the performance. This will help us to remove lot of variables.
going to have to see about that. Just started in a new position here so I hve to speak with the guy that engineered this network to see what we can do. The bad part is that this is a hospital so there really isn;t much downtime at all.
What about the local Datastore try that option and see let us know the results.