Problem with vmfs snapshot LUN

[sonvnicb]

Hi all.

Our system using two HP Blade and two Hitachi USPV storage (with truecopy replication) to provide DR solution with Vsphere 4. Now we have some problems with this solution:

1. When we splitted pair synchonize between two system and tried two rescan Datastorage on DR system there was an error message " Cannot change the Host configuration"in VCenter

2. Then, I tried to rescan and add storage directly to esxserver, it took very long time (about 10 minute) for each action, I tailed the vmkernel log file and found that esxserver have many message like:

Aug 9 09:18:27 esxserver-29 vmkernel: 23:17:13:27.144 cpu9:7149)ScsiDeviceIO: 7

47: Command 0x2a to device "naa.60060e8005b263000000b2630000001d" failed H:0x0 D

:0x2 P:0x0 Valid sense data: 0x7 0x22 0x0.

Anyone known this bug and how to fix it please help me.

Thanks

[ShaneWendel]

I believe this probably has something to do with having two LUNs with the same signature. I remember a post on either Duncan Epping or Mike DiPetrillo's blog about a situation similar to what you describe, but I can't seem to find it now. There is a command that you have to issue to resignature the disk.

I'll see if I can find it.

-

---

Shane Wendel, VCP

[ShaneWendel]

I was thinking of this: http://www.yellow-bricks.com/2009/05/04/unique-lun-ids/

Do you have both the primary lun and the replicated lun presented to the ESX servers at the same time? I guess the question is whether your SAN presents the LUN pair as a single unit or as two seperate LUNs. This might be confusing ESX.

-

---

Shane Wendel, VCP

http://fatalsync.wordpress.com

[sonvnicb]

Hi

We have two separate system on each site not same ESX. But two system is managed by one VC. I wondering this is problem?

More information: when I rescan storage with one esx ha 19 LUN, it takes about 10 minute to finished.

Anyone have solved problem, please help.