AntonyBarlow
Contributor
Contributor

Incorrectly Resignatured Luns

One of our ESX3.0.2 Servers has resignatured 2 of the LUN's attached to it. The other servers in the cluster can still access the LUN's under the correct name and the server with the problem access's them as snap-00000002-OldLunName.

The luns are presented via IBM's SVC. the path is different on the server with the problem. On the working servers the LUN paths are 1:0:0:1 and 1:0:1:1 and on the none working server the paths are 1:9:0:1 and 1:9:1:1.

Tags (2)
0 Kudos
5 Replies
Monoman
Enthusiast
Enthusiast

I ran into a similar problem. From what I remember you have to make sure the SVC presents all of the ESX hosts with same LUNs in the same order. The ESX hosts need all of the SCSI IDs to match and go to the same LUN for any shared storage.

I tried some work arounds when I had the issue but in the end resorted to creating completely new LUNs. Then I shuffled VMs around so I could delete the old LUNs. When I was done I had all of the VMs on the new LUNs and they LUNs matched for all of the ESX hosts.

I hope I didn't confuse you Smiley Happy

0 Kudos
Schorschi
Expert
Expert

Definite need to have 1) all LUNs presented via the same path policy, zoning, and LUN numbering. 2) Zoning once was recommended as one ESX per zone, but the new model recommended is all ESX hosts in the same zone that need to see the same storage. This requirement is also key to proper use of VCB.

0 Kudos
APlant
Enthusiast
Enthusiast

I agree with the zoning comments, but I suspect you will already have that correctly configured. I would also check that the LUN's are being re-presented with the same ID as they were originally presented.

Also take a look at the following parameters:

LVM.DisallowSnapshotLun=0 and LVM.EnableResignature=0

I have had to configure as above a few times when I've been configuring LUN mirroring. When I failed over to the secondary LUN I see a similar issue to you, once I change to the parameters listed above all was OK.

0 Kudos
Monoman
Enthusiast
Enthusiast

Also take a look at the following parameters:

LVM.DisallowSnapshotLun=0 and LVM.EnableResignature=0

I didn't have much luck when I changed those settings. I worked around it all by migrating all guests off of some of my hosts and recreating all new LUNs to migrate the guests back to.

0 Kudos
sbeaver
Leadership
Leadership

I have come across tis a couple of times and the only workaround that I have found was the same thing you did

Steve Beaver

VMTN Forum Moderator

*Virtualization is a journey, not a project.*

Steve Beaver
VMware Communities User Moderator
VMware vExpert 2009 - 2020
VMware NSX vExpert - 2019 - 2020
====
Co-Author of "VMware ESX Essentials in the Virtual Data Center"
(ISBN:1420070274) from Auerbach
Come check out my blog: [www.virtualizationpractice.com/blog|http://www.virtualizationpractice.com/blog/]
Come follow me on twitter http://www.twitter.com/sbeaver

**The Cloud is a journey, not a project.**
0 Kudos