Re: [Iscsitarget-devel] Abort Task / cmnd_abort

[unlisted <unl...@gm...>]

It's about 9 months later and I'm still seeing these during heavy iSCSI
usage.  Last week we actually had about 10 esx guests corrupted when this
happened.  Are there known issues with 0.4.16 and ESX 3.5 and multiple esx
boxes accessing the same vmfs volumes ?  Does anyone else use IET with
vmware and vmotion ?

Sam



On Tue, Aug 26, 2008 at 1:59 PM, unlisted <unl...@gm...> wrote:

>  Hey all,
>
> I've implemented an IET 0.4.16 target with 2 vmware ESX 3.5 servers
> accessing a shared lun.  Target is Debian with the 2.6.24 etchandahalf
> kernel.  Very basic ietd.conf configuration (modified sample) sharing up a
> volume sitting on top of LVM.  I've searched for an hour and came up with
> nothing on google or listservs.
>
> iscsi_trgt: cmnd_abort(1148) 3847ad2 1 0 42 512 0 0
> iscsi_trgt: Abort Task (01) issued on tid:1 lun:0 by sid:564049469047296
> (Unknown Task)
> iscsi_trgt: Abort Task Set (02) issued on tid:1 lun:0 by
> sid:564049469047296 (Function Complete)
> iscsi_trgt: Abort Task Set (02) issued on tid:1 lun:1 by
> sid:564049469047296 (Function Complete)
>
> Every few hours I see the above messages in syslog on that lun.  Is this
> normal activity?  Currently one of the ESX boxes is idle with nothing using
> the disk (just mounted) and I still see them, so it makes me wonder if its
> normal or not.
>
> Target iqn.1999-08.com.example:vmware-aangzuko
>         Lun 0
> Path=/dev/raid5/vmware-aangzuko,Type=blockio,ScsiId=VMDSK-20080818,ScsiSN=VMDSK-20080818,IOMode=wb
>         Lun 1
> Path=/dev/raid5/vmware-repository,Type=blockio,ScsiId=VMDSK-20080822,ScsiSN=VMDSK-20080822,IOMode=wb
>         Alias vmware-aangzuko
>         InitialR2T              No
>         ImmediateData           Yes
>         MaxRecvDataSegmentLength 8192
>         MaxXmitDataSegmentLength 65536
>         MaxBurstLength          262144
>         FirstBurstLength        65536
>         MaxOutstandingR2T       8
>         HeaderDigest            None
>         DataDigest              None
> Previously we we're using 2.6.18 kernel and didn't see these messages,
> although it's a complete different box where we are seeing this.  Single
> Intel 3xxx xeon...
>
> Thx,
> Sam
>