Networker

Re: [Networker] ADIC Scalar 10K and Windows storage node

2003-10-03 11:01:43
Subject: Re: [Networker] ADIC Scalar 10K and Windows storage node
From: Teresa Biehler <tpbsys AT RIT DOT EDU>
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Date: Fri, 3 Oct 2003 11:00:11 -0400
Davina,

I've looked at the SCSI reserve/release value in st.conf.  We currently
have it disabled at Legato's recommendation.  Also per Legato, since
Win2K does not have this parameter, it doesn't "respect" this setting
even if it is enabled on the Solaris side.

What you say about the QLOGIC and persistent bindings is true - it is
now supported.  Could you send me more information about the environment
you were using for testing (what type of hardware, configuration, etc.)?
If we start thinking about changing HBAs, I'll need more details to
justify the attempt.

Thanks.
Teresa

-----Original Message-----
From: Davina Treiber [mailto:treiber AT HOTPOP DOT COM] 
Sent: Friday, October 03, 2003 10:29 AM
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Subject: Re: [Networker] ADIC Scalar 10K and Windows storage node

On Fri, 3 Oct 2003 10:11:00 -0400, Teresa Biehler <tpbsys AT RIT DOT EDU>
wrote:

>We are zoning - I just forgot to mention that.  Here are the details.
>
>Our SAN is set up as two fabrics.  The Sc 10K has 4 drives attached to
>each SNC.  Each SNC is on one fabric.  The NW server is attached to
both
>fabrics and can see all 8 drives.  Each NW storage node is attached to
>both fabrics and can only access the 6 shared drives (4 from one SNC, 2
>on the other).
>
>Zoning is done by WWN.  The zones are set up from each server to the
>SNCs.  The NW server is included in two zones - one for each SNC.  Each
>storage node is included in two zones as well - one for each SNC.
>
>When we reboot one of the storage nodes while the NW server is doing a
>backup, the tape rewinds and overwrites the files that had existed
>previously on the tape.
>
>Do you have any suggestions?

Hi Teresa,

Some time ago I did some extensive testing of a SAN environment that
includes Solaris and Win2K servers. I was well aware of the potential
problems when rebooting a machine on the SAN.  I did a lot of tests with
rebooting during backups, and could not simulate the problem I had
anticipated. One thing you might look at is SCSI reserve/release
settings.
On Solaris this is configured in st.conf. In Windows there is no
reserve/release, it's just not a concept that Windows understands.

I was originally using QL2200F cards on Windows, but I abandoned these
because there was no support for persistent binding in the driver. I
hear
that this has now been introduced, and about time too. I gave up on
these
and installed Emulex cards, LP8000 and LP9002, and I liked these a lot
better. You could try Emulex cards, you may not have the reset problem -
I
didn't no matter how much I tried. And a couple of HBAs might be quite a
cheap solution compared to hours of problem-solving.

--
Note: To sign off this list, send a "signoff networker" command via
email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=

<Prev in Thread] Current Thread [Next in Thread>