We have VI3 Foundation edition (no VMotion) with three blade servers configured as host servers. Each host so far has about 5 VMs. All are pointed to an iSCSI SAN with each host pointed to it's own LUN on the SAN.
I was wondering what we need to do, if anything, to prepare for one of the physical servers to die. Since we're on shared storage we'd still be able to see the dead hosts VM's vmdk and other files, correct? Or will the separate LUN's prevent this? We can see them now via the VMware Infrastructure Client but will the dead servers VMs still show up after it dies?
If we can see the vmdk files and other vm files, can we re-create a VM from those after the fact or is there anything we need to do now before a server fails?
Thanks.
In the event of failure you can present the LUNS to another host and
the register those VMs on that host (From cli or Datastore Browser). Then boot the VMs and hopefully they all come back up. It would be a
little more complicated but that is about it.
In the event of failure you can present the LUNS to another host and
the register those VMs on that host (From cli or Datastore Browser). Then boot the VMs and hopefully they all come back up. It would be a
little more complicated but that is about it.
You can already present the LUN to another ESX and register the VM.
So you will be ready to just re-start them in case of host failure.
You can also create a simple script to start all your VM.
Andre
**if you found this or any other answer useful please consider allocating points for helpful or correct answers
Hey thanks. I wasn't familiar with the terms "presenting the LUN" and "registering the VM" but did searches and now I know what you're talking about. I'm good to go now. Thanks again.
"presenting the LUN"
Is just connect the same LUN on another ESX.
Usally there is some "storage side work" (in your case just authorize the second ESX).
"registering the VM"
Use Browse Datastore, go in VM folder, right click on vmx file and add to inventory.
Without VC you can have the same VM registered on more ESX (using shared storage).
Andre
**if you found this or any other answer useful please consider allocating points for helpful or correct answers