This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.
------_=_NextPart_001_01C56C2B.D1E79B9C
Content-Type: text/plain
In a previous life, I set up the following:
Windows 2k NBU servers single port HBAs 1 to tape SAN one to Disk SAN. We
would snapshot the data and mount the snapshot on the media servers and back
it up. Also our Exchange system had HP EVA disk and a separate HBA for Tape
and we made the exchange boxes media servers to do their own thing using the
exchange agent, so that it would not affect the other data backups. We used
customized scripts that would establish the mirror (EMC) at 1700 and break
the mirror at 2200 and mount the R2 (mirror) to various mount points on the
media servers. The scripts were intelligent enough to check if the R1/R2
were in a syncd state before splitting and mounting and would go into a wait
state if they were not and check again in 1 min intervals. The scripts also
had exstensive logging and email notifications based on various levels of
severity. Based on that it would notify the NOC and/or page the on-call for
the backup team.
The EMC disk SAN went through its own Mcdata switches and the tape SAN was
on its on Brocade switches.
It was a good setup and had fantastic through-put. Granted we were using
Micro$ofts diskpart (SLOW) util to do the mounts and unmounts since the
company did not want to spring for VTS Volume Manager. It would have made it
a lot easier. I'd highly recommend using volume manager for this.
All in all it was a very reliable setup and is still working to this day.
-----Original Message-----
From: veritas-bu-admin AT mailman.eng.auburn DOT edu
[mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu] On Behalf Of Paul
Keating
Sent: Wednesday, June 08, 2005 07:45 AM
To: veritas-bu AT mailman.eng.auburn DOT edu
Subject: RE: [Veritas-bu] Tape SAN + Data SAN
How many ports do you use for each machine?
We recently built up this Netbackup infrastructure with a view to SAN
attaching it in the future. Currently we only have a pair of V880s (each
with dual HBAs) and a pair of STK L700 with 4 FC LTO2 drives in each.
One V880 and one L700 live at each site. Each site has an 8 port FC switch.
(2 ports to the 2 HBAs on server, robot and 4 drives.) and the 8th port
connects to our Optera Metro 5200 which provides FC connectivity to our DR
site, where the mirror image lives (2nd FC switch, 2nd V880 and 2nd
L700)...the 2nd V880 is idle.....only receiving the NBU catalog using
Veritas Volume Replicator.
With our SAN install on deck, the switches will be replaced with Brocade
24k's on both sides, and everything will connect to these. The plan is for
every SAN connected device in our network to use dual HBAs managed with HDLM
(Hitachi dynamic link manager).
Our backup server already has dual HBAs for the tapes......
Are you using a single HBA for disk, and a single HBA for tapes? If so, does
that mean your servers are connected to only one FC switch? So you have
multiple single points of failure? Or are you using dual ported HBAs
connected to two switches? Therefore only the HBA is a SPOF??
I definitely would not want dual non connected fabrics. I've been pushing
this point here, as we're installing a 2nd 2gig FC pipe on our Opteras for
SAN replication traffic.....Some people are pushing to have the backup
traffic and replication traffic segregated....zones comepletely separately,
no talk between them......I've been saying that this keeps us from being
able to do "off host" backups.
How are most people doing backups of SAN attached volumes for servers that
are mouning disk directly from the SAN? Do you snapshot the volume, then
mount the snapshot on the media server, and back it up directly to tape? Or
do you install media servers on all the data servers with large amounts of
data, and use SSO?
Paul
> -----Original Message-----
> From: veritas-bu-admin AT mailman.eng.auburn DOT edu
> [mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu] On Behalf Of
> Mark.Donaldson AT cexp DOT com
> Sent: June 7, 2005 6:45 PM
> To: shazari AT gmail DOT com; veritas-bu AT mailman.eng.auburn DOT edu
> Subject: RE: [Veritas-bu] Tape SAN + Data SAN
>
>
> I mix the tapes & disks on existing SAN. I don't, however,
> mix tapes &
> disks on the HBA side on the servers. It's often recommended
> against by
> vendors because HBA settings that are good for disks are bad
> for tapes and
> vice versa. I run different zones for tape & disk.
>
> -M
_______________________________________________
Veritas-bu maillist - Veritas-bu AT mailman.eng.auburn DOT edu
http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu
------_=_NextPart_001_01C56C2B.D1E79B9C
Content-Type: text/html
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV=3D"Content-Type" CONTENT=3D"text/html; =
charset=3DUS-ASCII">
<META NAME=3D"Generator" CONTENT=3D"MS Exchange Server version =
5.5.2657.73">
<TITLE>RE: [Veritas-bu] Tape SAN + Data SAN</TITLE>
</HEAD>
<BODY>
<P><FONT SIZE=3D2>In a previous life, I set up the following:</FONT>
</P>
<P><FONT SIZE=3D2>Windows 2k NBU servers single port HBAs 1 to tape SAN =
one to Disk SAN. We would snapshot the data and mount the snapshot on =
the media servers and back it up. Also our Exchange system had HP EVA =
disk and a separate HBA for Tape and we made the exchange boxes media =
servers to do their own thing using the exchange agent, so that it =
would not affect the other data backups. We used customized scripts =
that would establish the mirror (EMC) at 1700 and break the mirror at =
2200 and mount the R2 (mirror) to various mount points on the media =
servers. The scripts were intelligent enough to check if the R1/R2 were =
in a syncd state before splitting and mounting and would go into a wait =
state if they were not and check again in 1 min intervals. The scripts =
also had exstensive logging and email notifications based on various =
levels of severity. Based on that it would notify the NOC and/or page =
the on-call for the backup team.</FONT></P>
<P><FONT SIZE=3D2>The EMC disk SAN went through its own Mcdata switches =
and the tape SAN was on its on Brocade switches.</FONT>
<BR><FONT SIZE=3D2>It was a good setup and had fantastic through-put. =
Granted we were using Micro$ofts diskpart (SLOW) util to do the mounts =
and unmounts since the company did not want to spring for VTS Volume =
Manager. It would have made it a lot easier. I'd highly recommend using =
volume manager for this.</FONT></P>
<P><FONT SIZE=3D2>All in all it was a very reliable setup and is still =
working to this day.</FONT>
</P>
<P><FONT SIZE=3D2>-----Original Message-----</FONT>
<BR><FONT SIZE=3D2>From: veritas-bu-admin AT mailman.eng.auburn DOT edu [<A =
HREF=3D"mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu">mailto:veritas-b=
u-admin AT mailman.eng.auburn DOT edu</A>] On Behalf Of Paul =
Keating</FONT></P>
<P><FONT SIZE=3D2>Sent: Wednesday, June 08, 2005 07:45 AM</FONT>
<BR><FONT SIZE=3D2>To: veritas-bu AT mailman.eng.auburn DOT edu</FONT>
<BR><FONT SIZE=3D2>Subject: RE: [Veritas-bu] Tape SAN + Data SAN</FONT>
</P>
<BR>
<P><FONT SIZE=3D2>How many ports do you use for each machine?</FONT>
</P>
<P><FONT SIZE=3D2>We recently built up this Netbackup infrastructure =
with a view to SAN attaching it in the future. Currently we only have a =
pair of V880s (each with dual HBAs) and a pair of STK L700 with 4 FC =
LTO2 drives in each.</FONT></P>
<P><FONT SIZE=3D2>One V880 and one L700 live at each site. Each site =
has an 8 port FC switch. (2 ports to the 2 HBAs on server, robot and 4 =
drives.) and the 8th port connects to our Optera Metro 5200 which =
provides FC connectivity to our DR site, where the mirror image lives =
(2nd FC switch, 2nd V880 and 2nd L700)...the 2nd V880 is idle.....only =
receiving the NBU catalog using Veritas Volume Replicator.</FONT></P>
<P><FONT SIZE=3D2>With our SAN install on deck, the switches will be =
replaced with Brocade 24k's on both sides, and everything will connect =
to these. The plan is for every SAN connected device in our network to =
use dual HBAs managed with HDLM (Hitachi dynamic link =
manager).</FONT></P>
<P><FONT SIZE=3D2>Our backup server already has dual HBAs for the =
tapes......</FONT>
</P>
<P><FONT SIZE=3D2>Are you using a single HBA for disk, and a single HBA =
for tapes? If so, does that mean your servers are connected to only one =
FC switch? So you have multiple single points of failure? Or are you =
using dual ported HBAs connected to two switches? Therefore only the =
HBA is a SPOF??</FONT></P>
<P><FONT SIZE=3D2>I definitely would not want dual non connected =
fabrics. I've been pushing this point here, as we're installing a 2nd =
2gig FC pipe on our Opteras for SAN replication traffic.....Some people =
are pushing to have the backup traffic and replication traffic =
segregated....zones comepletely separately, no talk between =
them......I've been saying that this keeps us from being able to do =
"off host" backups.</FONT></P>
<P><FONT SIZE=3D2>How are most people doing backups of SAN attached =
volumes for servers that are mouning disk directly from the SAN? Do you =
snapshot the volume, then mount the snapshot on the media server, and =
back it up directly to tape? Or do you install media servers on all the =
data servers with large amounts of data, and use SSO?</FONT></P>
<P><FONT SIZE=3D2>Paul</FONT>
</P>
<BR>
<P><FONT SIZE=3D2>> -----Original Message-----</FONT>
<BR><FONT SIZE=3D2>> From: =
veritas-bu-admin AT mailman.eng.auburn DOT edu</FONT>
<BR><FONT SIZE=3D2>> [<A =
HREF=3D"mailto:veritas-bu-admin AT mailman.eng.auburn DOT edu">mailto:veritas-b=
u-admin AT mailman.eng.auburn DOT edu</A>] On Behalf Of </FONT>
<BR><FONT SIZE=3D2>> Mark.Donaldson AT cexp DOT com</FONT>
<BR><FONT SIZE=3D2>> Sent: June 7, 2005 6:45 PM</FONT>
<BR><FONT SIZE=3D2>> To: shazari AT gmail DOT com; =
veritas-bu AT mailman.eng.auburn DOT edu</FONT>
<BR><FONT SIZE=3D2>> Subject: RE: [Veritas-bu] Tape SAN + Data =
SAN</FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> I mix the tapes & disks on existing =
SAN. I don't, however,</FONT>
<BR><FONT SIZE=3D2>> mix tapes &</FONT>
<BR><FONT SIZE=3D2>> disks on the HBA side on the servers. =
It's often recommended </FONT>
<BR><FONT SIZE=3D2>> against by</FONT>
<BR><FONT SIZE=3D2>> vendors because HBA settings that are good for =
disks are bad </FONT>
<BR><FONT SIZE=3D2>> for tapes and</FONT>
<BR><FONT SIZE=3D2>> vice versa. I run different zones for =
tape & disk.</FONT>
<BR><FONT SIZE=3D2>> </FONT>
<BR><FONT SIZE=3D2>> -M</FONT>
</P>
<P><FONT =
SIZE=3D2>_______________________________________________</FONT>
<BR><FONT SIZE=3D2>Veritas-bu maillist - =
Veritas-bu AT mailman.eng.auburn DOT edu <A =
HREF=3D"http://mailman.eng.auburn.edu/mailman/listinfo/veritas-bu"; =
TARGET=3D"_blank">http://mailman.eng.auburn.edu/mailman/listinfo/veritas=
-bu</A></FONT>
</P>
</BODY>
</HTML>
------_=_NextPart_001_01C56C2B.D1E79B9C--
|
