We have been backing up Microsoft Clusters for sometime now.
I have each node separated out: c:\, d:\, SYSTEM STATE:\ etc...
However (I never tried) if you were to do an ALL on the nodes and say
everything is on node 1 then you would get all the drives and if/when it
should fail over to node 2 you would get them all there too?? If that
were the case then you wouldn't need cluster license? Never tried, I am
satisfied with the way we have the backups setup...as follows:
For the 'drives' we have had to set up a virtual client name and list
them individually. Also in the clients resources on the server you need
to put each node in Remote Access (system@node1, system@node2, etc).
If you do end up having to setup all the drives individually you can
list all the possible drive letters in the virtual client. It will
backup the drives that are available and the ones that aren't set up it
will come back and say "no such file or directory" When/if they put a
new drive out there, when the backup runs it should see it and do a FULL
the first time. Just will need to check your logs.
One of my clients a directory will fill up and they would create a new
one and this is how I could tell when the one directory was full and
they started on a new one, otherwise I would be backing up directory 1
and things were written to directory 2 and I wouldn't know. I do this
with directories, and assume it would work similar with drives?
Kathleen Scalise
-----Original Message-----
From: Legato NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU]
On Behalf Of Tim Verbois
Sent: Friday, February 03, 2006 3:45 AM
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Subject: Re: [Networker] MSCS cluster backups
I think you need to list the drives for the physical nodes as well. We
have done it that way, advised by Sun. A Sun Solaris cluster is
different, there you can assign the slices to the different hosts
(physical and virtual) and he will only backup for that client what the
client ownes when you use All as saveset.
Microsoft is not that far advanced :-P, just kidding (wouldn't wanna
step on any toes :-P)
Tim
Teresa Biehler wrote:
>We had this problem when we were setting up a Windows cluster. From
>what I remember, it says in the admin guide that "all" should work for
>the physical nodes, but that the virtual node needs to have the drives
>listed out. I wish it were otherwise too.
>
>>From the service call we had opened, I vaguely remember that we had to
>add "-c client_name" to the save command. I believe Legato also wanted
>us to add the physical hosts and virtual hosts to the local hosts file
>on all systems of the cluster.
>
>Teresa
>
>-----Original Message-----
>From: Legato NetWorker discussion
>[mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU]
>On Behalf Of Darren Dunham
>Sent: Thursday, February 02, 2006 5:38 PM
>To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
>Subject: [Networker] MSCS cluster backups
>
>Is anyone doing a lot of MSCS backups? I have some questions.
>
>If at all possible, I'd like to avoid hard-coding drive letters in the
>save sets on the server. I'm afraid of the cluster
>administrators/owners adding a drive and not telling the backup folks.
>
>In some testing I've found that putting "All" in both the physical node
>backup and the virtual host backup works some, but not all, of the
>time. Unfortunatly, I can't find what's different between when it
>works and when it doesn't.
>
>I have one cluster with 2 vhosts currently online on one node. When I
>hit one vhost, a probe returns only the drive allocated to it. When I
>hit the other vhost, the probe returns all of the other drives (both
>that vhost shared and local to the node). This is the same list I get
>when I hit the physical node. :-(
>
>Unlike VCS on unix, I'm finding it difficult to tease out where the
>savefs probe is looking for stuff. I called support, but they're
>telling me that what I want to do is not recommended. They'd rather I
>go and hardcode the save sets and can't give me much information on
>what might be wrong in my cluster or networker setup.
>
>Do I need to give up this idea that I should be able to use "All" here?
>I'm doing that successfully in another cluster and just assumed it was
>correct and supported.
>
>Current clients and server are 7.1.1. Any help for how to decipher
>what's going on here is appreciated.
>
>Thanks!
>
>
>
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems wit this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or via RSS at
http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
--------------------------------------------------------
If you are not an intended recipient of this e-mail, please notify the sender,
delete it and do not read, act upon, print, disclose, copy, retain or
redistribute it. Click here for important additional terms relating to this
e-mail. http://www.ml.com/email_terms/
--------------------------------------------------------
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the
body of the email. Please write to networker-request AT listserv.temple DOT edu
if you have any problems
wit this list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
|
