I can now get the deployment to finish. I ran snoop on the OpenSolaris box
and had a warnings about MTU size:

(warning) packet length greater than MTU in buffer offset

I fixed this by enabling jumbo frames in the config file for my e1000 NICs.

2008/12/12 Peter Rosenthal <[email protected]>

> Hello,
>
> I am using iscsitgtd via zfs shareiscsi=on. I am attempting to use it to
> store an ESX VMFS volume and I'm having a few difficulties. I have no
> problems with the initiator connecting to the target, I can format the
> device and everything looks good. However, when I attempted to deploy a 4GB
> VM to the store it failed around 94% with 'time out' and there are several
> messages in the ESX log:
>
> Dec 12 11:06:27 vmlab03 vmkernel: 0:08:14:37.142 cpu1:1031)Fil3: 5004:
> WRITE error 0xbad00e5
>
> Dec 12 11:06:59 vmlab03 vmkernel: 0:08:15:09.146 cpu1:1031)Fil3: 5004:
> WRITE error 0xbad00e5
> Dec 12 11:06:59 vmlab03 vmkernel: 0:08:15:09.146 cpu1:1031)Fil3: 9875: Max
> timeout retries exceeded for caller 0x8e42bb (status 'Timeout')
>
> Dec 12 11:07:09 vmlab03 vmkernel: 0:08:15:19.793 cpu0:1024)VMNIX: VMFS:
> 245: timed out
> Dec 12 11:07:20 vmlab03 vmkernel: 0:08:15:30.800 cpu1:1037)FS3: 4831:
> Reclaimed timed out heartbeat [HB state abcdef02 offset 4094976 gen 23 stamp
> 29695268365 uuid 49424
> 253-77842959-564d-000a5e01e1db jrnl <FB 231603> drv 4.31]
>
> In many ways it seems to be like:
> http://opensolaris.org/jive/thread.jspa?messageID=310330
>
> But I am not using broadcom, nor do I have I/O AT module listed when I do
> modinfo.
>
> Anyone have any suggestions on how I should proceed with debugging this
> from the OpenSolaris side? I am running 2008.11
>
>
_______________________________________________
storage-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/storage-discuss

Reply via email to