On Tue, May 29, 2012 at 5:46 PM, Richard Elling
<[email protected]> wrote:
> idea at the bottom...
>
> On May 29, 2012, at 12:56 PM, Jason Cox wrote:
>
>> Let me start by saying that I am very new to OpenIndiana and Solaris
>> 10/11 in general. I normally deal with Red Hat Linux. I wanted to use
>> OI for ZFS support for a vmware shared storage server to mount LUNs on
>> my SAN.
>>
>> Setup:
>>
>> 2 servers with multiple fiber-channel connections directly connected
>> to my SAN (no SAN switch).
>>
>>
>> Situation:
>>
>> I have multipath working and I create the zpool with no problem using
>> the multipath disk device.
>> ---
>> [email protected]:~# zpool create lun00 
>> c2t60060E80104B8F6004F327FE00000000d0
>> [email protected]:~# zpool status lun00
>>  pool: lun00
>> state: ONLINE
>>  scan: none requested
>> config:
>>
>>        NAME                                     STATE     READ WRITE CKSUM
>>        lun00                                    ONLINE       0     0     0
>>          c2t60060E80104B8F6004F327FE00000000d0  ONLINE       0     0     0
>>
>> errors: No known data errors
>> ---
>>
>> Now I can export the pool from nfs01 to nfs02 with no problem..
>>
>> ---
>> [email protected]:~# zpool status lun00
>>  pool: lun00
>> state: ONLINE
>>  scan: none requested
>> config:
>>
>>        NAME                                     STATE     READ WRITE CKSUM
>>        lun00                                    ONLINE       0     0     0
>>          c2t60060E80104B8F6004F327FE00000000d0  ONLINE       0     0     0
>>
>> errors: No known data errors
>> ---
>>
>> The issue comes up when I then export the pool off nfs02 and try to
>> import it again back on nsa01.
>>
>> ---
>> [email protected]:~# zpool import lun00
>> Assertion failed: rn->rn_nozpool == B_FALSE, file
>> ../common/libzfs_import.c, line 1093, function zpool_open_func
>> Abort (core dumped)
>> ---
>>
>> No matter how many times I try to import the pool on 01 I have this
>> issue. Both servers are running the same version of OI and all the
>> same updates. They are also the same servers purchased and spec'ed at
>> the same time for this project.
>>
>>
>> Any guidance would be appreciated.
>
> This can occur if the disk label does not look the same from both systems.
>        prtvtoc /dev/rdsk/c2t60060E80104B8F6004F327FE00000000d0s0
> and compare
>  -- richard
>
> --
> ZFS Performance and Training
> [email protected]
> +1-760-896-4422
>
>
>
> _______________________________________________
> OpenIndiana-discuss mailing list
> [email protected]
> http://openindiana.org/mailman/listinfo/openindiana-discuss


I have an update on this... After talking to a friend who deals with
Solaris all the time, he was thinking it was a bug in EFI. Based on
what Richard was thinking too, I took a smaller LUN that was below 2TB
and changed the label from EFI to SMI. It works now. I can move the
LUN back and forth between the two servers. So I guess I either redo
all my LUNs to support SMI and let ZFS RAID them together or I wait
and see if a patch comes out for this issue.

-- 
Jason Cox

_______________________________________________
OpenIndiana-discuss mailing list
[email protected]
http://openindiana.org/mailman/listinfo/openindiana-discuss

Reply via email to