storage problem please help

[mtsm]

Hi people , i recently added 4 new hosts to my vmware cluster...everything is okay , the two hba from the 4 new hosts can see the same luns... but when i do a refresh or reboot the filesystems are not recognized , a friend of mine told me to change a parameter in advanced settings called enable lvm resignuturing to 1 , did that...and the things got worst

now i can see the filesystems but with crazy names , like SAN28 became snap-3340238-SAN28 in the new server...when i try to rename it , it says "SAN 28 already exists" what can i do? im really afraid of doing anything that will fuckup my filesystems

anyone have ideas?

[Dave_Mishchenko]

You're going to want to call VMware support on this and have them look at your system. The EnableResignature option should be used when you want to attach to snapshots or replicas of VMFS datastores to the same set of servers.

In your case it should have been sufficient to just rescan for storage and new hosts should have recognized the existing datastores. It is likely the case, that the new hosts were not using the same LUN ID for your shared storage as the existing hosts. Hence, the servers did not recongnize them as existing datastores, but rather as snapshots (hence snap as part of the datastore name) when you used the resignature option.

Page 118 covers the resignature option - http://www.vmware.com/pdf/vi3_35/esx_3/r35/vi3_35_25_san_cfg.pdf

[espi3030]

The very first thing you should do right now is try to make a backup of the VM's, or vmotion them to another host. Two weeks ago I was EXACTLY were you are now, my best advice is make a backup. And what ever you do, do not reboot the ESX host thinking that will fix it, because it wont.

Like Dave said, call VMware and they should be able to help you. There is a way out of this as long as you don't panic, I did and lost (4) important VM's.

[mtsm]

now im really scared...have only one hosts with all my vms running...the other ones i rebooted , some of them wont recognize the filesystems , some of them have the filesystems with snap-3ae1a-SAN28 .....and the worst , out support contract has expired , we are trying to correct this contract thing...

im gonna wait with only one host running until there

[Dave_Mishchenko]

You should be OK with this. It's just best to have a single person stick with you through the entire process. You'll want to check your SAN setup especially how you've assigned LUN IDs. As mentioned the LUN ID for a specific LUN has to be the same accross all hosts.

[espi3030]

I did not mean to scare you, at least you can still see something. In my case I could not see the storage LUNs at all and that is why I could not backup my VM's, you are in alot better shape than I was. What I should of done before I rebooted (thus losing SAN visibility) is have a new LUN presented so I can make a back up of my VM's, either by relocating the VM's to the new LUN or via backup script, then powering them up on a new host.