This is an old post i made on the VMware communities earlier this year, i am re-posting it as i don’t want it to get buried, given the amount of post they get in a day :D.
So I decided to post an observation i made during a non standard ESXI setup (couldn’t find anything about it so i am making a post).
This is not a fault or a question more of a discussion.
The circumstance below are due to the limitation of having only one server in a remote site.
What i noticed and i am pretty sure this is not supported or recommended, is that when you have a virtual machine on an esxi host that presents (as a datastore) a NFS full of ISO to VMWare (the esxi host) the machine hosting the NFS cannot use the ISO.
What i mean by the above is if you try and mount an ISO from the NFS datastore on the virtual machine hosting them it will lock up. Now if you try and mount it on any other Virtual machine it’s fine no problem.
The bellow diagram show our setup.
I assume it has something to do with the reconfiguration of the virtual machine when you mount the ISO, maybe a microsecond downtime for it to mount and due to this downtime the ISO presented becomes unavailable and in turn locks up the mount, it’s worth noting when i say it locks up the entire virtual machine locks up. Eventually vmware does pop up a message and ask if i should cancel the mount as it noticed that it’s causing problems. Maybe someone with better understating than me can explain why this behavior occurs.
Now i know there are other ways of doing this like putting the images directly on the datastore but our situation requires that ISO s are synced on regular bases so we use DFS to sync from main site and NFS(usually located on different server) to present to ESXI servers.