• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Vcenter Server status hangs while making a new VM

A

AMD_Gamer

Fully [H]
Joined
Jan 20, 2002
Messages
18,287
Ran into an interesting problem today. I tried to create a new VM on one of my hosts and the progress % bar would keep hanging at a certain point and would never finish. In the performance monitor you can see still the network, disk, etc.. still working. I eventually restarted my server running Vcenter Server and when it came back up it appears to have successfully made the VM. I found a few things on google that says this is probably an error with how Vcenter is reporting from the host? Has anyone run into this problem before I have a lot of work to do on my hosts and this would be a PITA and not how I want to deliver it to my boss.
 
What percentage is it hanging at? Instead of restarting the vcenter server, does just restarting the VMware VirtualCenter Server service fix the issue as well? Is it only happening on that one host or does it happen with all of them?
 
ND40oz said:
What percentage is it hanging at? Instead of restarting the vcenter server, does just restarting the VMware VirtualCenter Server service fix the issue as well? Is it only happening on that one host or does it happen with all of them?
Click to expand...

I was just thinking about checking the other host tomorrow. I was doing this right before I left work. One time it hanged at 13% the next time 26%.

how do i restart the VirtualCenter service? I will try this tomorrow http://kb.vmware.com/selfservice/mi...nguage=en_US&cmd=displayKC&externalId=1003895
 
Another problem is that both of my hosts show the yellow warning triangle in Vcenter Server but i do not see any problems with them?:confused:
 
I have the same issue with my ESXi host at home. Sometimes vcenter hangs, especially when making new VM's and I have to restart the server box... weird stuff.
 
I switched from iSCSI to NFS - ESXi seemed a lot more finicky about the former...
 
This only happens on one of the hosts. Creating the virtual machine takes forever and hangs at a certain % everytime. I tried putting a VM on the iSCSI and local datastore. On my second ESXi host creating a VM takes seconds. What could cause this?:confused:
 
Your DNS settings the same on both hosts? You have forward and reverse entries in for both hosts?
 
ok i added the hosts in the forward lookup zones. Will see if anything changes,
 
Ok i just figured it out and not sure if this is normal but this slow creation of the VM's only happens when i select the "Support Clustering Features such as Fault Tolerance" but wait.......i now feel like an idiot because it clearly says underneath it that this will increase the time it takes to create the VM lol. So far today I have not had the problem of it hanging since i restarted the Vcenter server so assume the service restart fixes that problems.


Now here is a question. If you created a VM without the support for clustering can you still use that VM with clustering features in the future or do you have to have the clustering option selected from the creation of the VM?
 
Last edited:
LOL

The reason it is taking so long when you choose fault tolerance is because it is creating an eagerzeroed thick .vmdk. It will create a thick disk, then it will "write zeros" to the whole thing before the provisioning is finished. If you don't select support fault tolerance features or clustering, then it just uses a zeroed thick .vmdk which writes zeros on write, but it isn't "prezeroed" at create time.

Doing the full zeroing at creation improves performance when writing to non-stale data areas of the disk, since it doesn't have to zero out first. Its not much of an improvement, but it is there. So I guess the question is... why use thick disks, and are you patient enough for that small gain ? :).

In order to have the clustering be "supported" you need to use the eagerzeroed thick disks -- yes. I've never tried it without. I imagine it would work, but that isn't all that smart of a move.
 
defuseme2k said:
LOL

The reason it is taking so long when you choose fault tolerance is because it is creating an eagerzeroed thick .vmdk. It will create a thick disk, then it will "write zeros" to the whole thing before the provisioning is finished. If you don't select support fault tolerance features or clustering, then it just uses a zeroed thick .vmdk which writes zeros on write, but it isn't "prezeroed" at create time.

Doing the full zeroing at creation improves performance when writing to non-stale data areas of the disk, since it doesn't have to zero out first. Its not much of an improvement, but it is there. So I guess the question is... why use thick disks, and are you patient enough for that small gain ? :).

In order to have the clustering be "supported" you need to use the eagerzeroed thick disks -- yes. I've never tried it without. I imagine it would work, but that isn't all that smart of a move.
Click to expand...

Multiwriter enabled but non-zeroed disks will keep the vm from booting :)
 
So from my understanding is that if you do not select this option during the VM creation it will do this when you add it to a cluster? You don't need to do it from the start if you want to use it for clustering and HA in the future?
 
AMD_Gamer said:
So from my understanding is that if you do not select this option during the VM creation it will do this when you add it to a cluster? You don't need to do it from the start if you want to use it for clustering and HA in the future?
Click to expand...

What type of disk you use has nothing to do with ESXi clustering or HA. Enabling HA or Clustering or both will do nothing to the disk, and disks do not need to be eagerzero to be in a HA Cluster.

It has to do with FT, which is a feature of clustering and HA. The disks DO NOT have to be eagerzero thick disks if doing HA and clustering, they DO if you want to use FT, which is a feature of HA, but not HA.

HA = Any style of disk(thin, thick, eagerzero thick)
CLUSTER = Any style of disk(thin, thick, eagerzero thick)
HA CLUSTER = Any style of disk(thin, thick, eagerzero thick)
FT = Must be eagerzero

In addition, FT is a PER VM setting, if you use FT only THAT VM WHICH USES IT requires eagearzero thick disks. All the other disks in the HA Cluster can be thin or w/e else. You CAN change/clone/remake/etc(more than one way to do it) the disk from any style(thin, etc) into eagerzero, enabling FT however will NOT do it automatically for you. It will simply fail to start the VM until you modify the disk to eagerzero.
 
Last edited:
It also has to do with MSCS clustering, aka multiwriter mode, using shared VMDKs. Without it, goofy things happen, hence why it's blocked.
 
You must log in or register to reply here.
Back
Top