Re: [Scst-devel] Scsi command errors

[Ming Z. <mi...@el...>]

On Wed, 2005-11-09 at 11:02 +0100, Johan Kragsterman wrote:
> Hi! I=C2=B4m new to this list, but known by some of you from iet list. =
First I=20
> want to thank Vlad for his efforts, this is a nice piece of software, a=
nd=20
> we hope that we will be able to contribute to the developement of it. W=
e=20
> are going to use scst in a bigger project, that involves iet, evms, and=
=20
> some other software. We have a professional storage setup, with many=20
> oportunities. At the moment, we have setup like this: First a workbench=
 as=20
> a host, running Win2k, with a QLA 2202 HBA connected to a SANbox switch=
.=20
> The switch is connected to QLA 2200 target HBA on Storior, our experime=
ntal=20
> storage appliance, which runs Lineox Enterprise Linux 4.0(a Red Hat clo=
ne),=20
> kernel 2.6.14. Storior is then connected through an LSI 7202 initiator =
HBA=20
> to another SANbox switch, and that switch is finally connected to an IB=
M Fc=20
> ProFibre 1100J JBOD storage array, with 5 st318452fc seagate fc drives.=
=20
> Through scst we exported two drives ask disk devices, and they showed u=
p=20
> nicely in the workbench HBA bios setup list of devices. When we went in=
 to=20
> the HBA setup, we where able to verify data and media. It was also=20
> functioning very well to make volumes and format them in windows, and a=
fter=20
> that to move data around. No problems with some quick tests. But when w=
e=20
> tried to low level format, through the workbench HBA setup tool, we got=
=20
> scsi command errors, like:
> scsi operation code: 04

format_unit

> scsi sense key: 06

unit attention

> scsi additional sense code: 29
> scsi additional sense code qualifier: 00


29/00  DTLPWROMAEBKVF  POWER ON, RESET, OR BUS DEVICE RESET OCCURRED


>=20
> After this, the drives were corrupt, and it was impossible to access th=
em=20
> from the workbench, and all the way to the LSI 7202 HBA on Storior. In =
the=20
> Storior LSI 7202 HBA setup tool, you could not access the drives with t=
he=20
> verifying tool, so you have to low level format them, which went allrig=
ht.=20
> Now the drives are functional again.
>=20
> Of coarse we want to know what was happening here. A theory might be th=
at=20
> the QLA HBA didn=C2=B4t work very nice together with the LSI HBA, when =
it sended=20
> the low level format command. JK

format_unit might need a very long scsi timeout?

ming

>=20
>=20
>=20
>=20
> -------------------------------------------------------
> SF.Net email is sponsored by:
> Tame your development challenges with Apache's Geronimo App Server. Dow=
nload
> it for free - -and be entered to win a 42" plasma tv or your very own
> Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
> _______________________________________________
> Scst-devel mailing list
> Scs...@li...
> https://lists.sourceforge.net/lists/listinfo/scst-devel