On Fri, Jul 23, 1999 at 05:37:00AM +0000, Steve Underwood wrote:
> Willy Tarreau wrote:
>=20
> > Hi !
> >
> > You could physically share the same SCSI disks array :
> >
> > 2 hosts connected to the same disk array, only one mounts the FS R/W, t=
he
> > other one mounts them R/O (or does not mount them), and when a problem =
is
> > detected, the backup mounts the FS R/W (in case of some crashes, you'll
> > have to do FSCK before that), and does the IPAT.
> >
> > Anyway, this only prevents from software or host hardware crash, but not
> > from file corruption nor disk problem in general.
> >
> > Willy
>=20
> Try this with most SCSI card/driver combinations and you will have troubl=
e.
> The bus reset performed by the second SCSI card to start up will screw up=
the
> operation of the first SCSI card. IF you use some sequencing trick to get
> things to start up you still cannot reboot one backup machine without
> screwing up the operation of the other.
>=20
> With some SCSI cards a specially adapted driver could get around this
> problem, but with other cards the hardware prevents this from ever working
> reliably.
I can not confirm the problem with SCSI resets.
Basically, a driver has to detect, when a SCSI reset occurs.
Ii has to tell the mid-layer that all commands already sent were aborted
(DID_RESET) and reset it's negotiation params (Sync, Wide). That's all,
basically.=20
I've operated a SCSI bus with several controllers without problems, BTw. The
combination was a NCR53c875 (DC390F) and a AM53C974 (DC390).
I see another problem: The box which mounts the FS read-only will have
inconcsistent buffers and you'll probably see lots of fs warnings. The
reasons is buffering: The data and metadata on disk has been updated by the
RW box, but the RPO box has still outdated buffers. Now, when some buffers
are read in, you get inconsistent states. So your RO box should bypass the
buffer cache. (Test it and use flushb for discarding the buffers.)
Regards,
--=20
Kurt Garloff <garloff@suse.de> SuSE GmbH, N=FCrnberg, FRG
Linux kernel development; SCSI drivers: tmscsim(DC390), DC395
--Dzs2zDY0zgkG72+7
Content-Type: application/pgp-signature
-----BEGIN PGP SIGNATURE-----
Version: 2.6.3in
iQCVAwUBN5hS1RaQN/7O/JIVAQEDYwP+KjZJoVdzetZhJV3oNbPmY+bWrhE5Zkp7
Wa8MRi4dk+ytrwR6sBoGB9xJztyj7qB9IUYja+wYV0RbRb5sq+KHRnmh1lTBPA6l
sQFJilRB+I8CI4X7d4LzqD7ZDHeSdtCdwHzSKrfKzWnNCXCN7RA1UdqhSWwThXox
RipNjyIkiAY=
=eKHc
-----END PGP SIGNATURE-----
--Dzs2zDY0zgkG72+7--
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/