would cycling the iscsi service to try to run an update wipe data?

Status
Not open for further replies.

rp503

Cadet
Joined
Nov 16, 2012
Messages
1
I've been running an esxi server and freenas box over the summer, other then the iscsi seeming to run extremely slow I haven't had any issues.

I started looking into the slowness and in the process decided to try and run the update from 8.0.3 release x64 to the latest.

I turned off the iscsi process and tried to run the updater, the update didn't happen. Oddly it didn't see any drive to store the temp files to so I cancelled the update and turned the iscsi back on.

last night I noticed that a webservice that is running on the VM I have on the esxi wasn't accepting connections. This morning I log into the esxi server to find the VM is down. Eventually I realized that the entire VM folder and its files is missing from the freenas scsi container.

Is it possible that recycing the iscsi service killed the data? What else would I have had to do to have the iscsi container delete its data?

The freenas is seeing the drives as 1 logical drive raided via motherboard controller.
 

survive

Behold the Wumpus
Moderator
Joined
May 28, 2011
Messages
875
Hi rp503,

I can only speak about what I've seen, but I've done terrible, terrible things to my iscsi LUN's presented to ESXi without any ill effects. I can confidently say that I haven't been able to lose any data when I has a NIC incompatibility that would kernel panic the box, so I'd be pretty surprised if shutting down the iscsi service would trash your VMFS volume.

-Will
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,680
Shouldn't. However, is your disk full? What I *have* seen is for ESXi's vmfs to get in a state where it is trying to write a block to disk, and there isn't space on the FreeNAS box to allocate a block, and it fails, and then a whole bunch of unwritten blocks pile up on ESXi, and at some point it calls it quits and the datastore is suddenly "empty", and if you're unlucky you might lose some data to corruption. The hopefully good news is that fixing the out-of-space condition and rebooting ESXi is likely to bring back what was there, though again you might lose some stuff to corruption, so careful fsck's are advised.
 
Status
Not open for further replies.
Top