I had the following virtual environment set up which was running for 2 years.
Two dedicated ESXi which are connected via fibre and iSCSI to TrueNAS.
Suddently, the cpu usage of the ESXi rised up to 100% and I found something like “no valid partition table” in the logs.
I connected an linux (pve with debian), installed the vmfs-tools and established the iSCSI connection to TrueNAS.
Only one small MS partition was found for at least some MB. It seems that everything was gone ;-(
I’ve two questions:
How can this happen and what can raise such a behavior (root cause)?
Is it possible to re-create the partition table, so that I can mount the vmfs and restore the files?
I can’t answer your questions because I don’t know the formatting of a VMFS file but I hope someone can help you. However here are a few questions you can answer that may help someone help you. I do hope you had a backup of your data.
What version of ESXi?
What version of TrueNAS?
Define the physical makeup of the TrueNAS server.
I don’t want to read into this but if you are saying that you were running TrueNAS on ESXi (not a problem at all), and the storage was a fake drive (VMFS) then that is absolutely a thing not to do. I only do that with the boot drive, never with the data I’m storing. You pass through the controller for the drives and then all the drives are allowed to operate as designed.
Thanks for the reply but actually, I don’t have access to the system and I can provide the requested information next week ;-(
(should be esxi 7 and tuenas core; truenas is using zfs)
I would like to re-use the zfs for two/three pve (Proxmox) server instead of ESXi. And, I would like to use zfs snapshots as well on the TrueNAS - that’s one of the advantages of zfs which should be used. On pve, i will connected to truenas via iscsi again and use lvm over the zfs provided by truenas.
