[Stgt-devel] Stability problem

[FUJITA Tomonori]

On Mon, 26 May 2008 14:46:35 +0200
Tomasz Chmielewski <mangoo at wpkg.org> wrote:

> FUJITA Tomonori schrieb:
> > On Thu, 15 May 2008 15:37:02 +0200
> > Konrad Bechler <kbechler at bakertillypoland.eu> wrote:
> > 
> >> Hello,
> >>
> >> I've got small, testing environment. I use Fedora 7 with 
> >> 2.6.23.15-80.fc7 as a iSCSI target and Windows XP box with "Microsoft 
> >> iSCSI Initiator" as a initiator.
> >> I configured stgt with only one target and one LUN:
> >> Target 1: iqn.2008.05.eu.tgc.net:storage.disk1
> >>      System information:
> >>          Driver: iscsi
> >>          State: ready
> >>      I_T nexus information:
> >>      LUN information:
> >>          LUN: 0
> >>              Type: controller
> >>              SCSI ID: deadbeaf1:0
> >>              SCSI SN: beaf10
> >>              Size: 0 MB
> >>              Online: Yes
> >>              Removable media: No
> >>              Backing store: No backing store
> >>          LUN: 1
> >>              Type: disk
> >>              SCSI ID: deadbeaf1:1
> >>              SCSI SN: beaf11
> >>              Size: 10486 MB
> >>              Online: Yes
> >>              Removable media: No
> >>              Backing store: /data/disk1
> >>      Account information:
> >>      ACL information:
> >>          ALL
> >>
> >>
> >> I can connect to this target. But when trying to format such drive, I've 
> >> got message, that formatting exited with error. When "quick format" box 
> >> is checked, tgtd aborts.
> > 
> > Last week, Konrad helped me to figure out what's the problem. Seems
> > that the MODE_SENSE data corruption leads to this, which Ronnie
> > reported in another thread.
> 
> Does it only affect some Windows initiator? Or can it possibly affect 
> open-iscsi as well?

Theoretically, yes. This affects any initiator implementations. I'm
not sure if it affects open-iscsi in reality.