Robert,
u can use Dirk Verhoeff suggestion.
on the client configuration for normal backup, just tick on no_index_save
but end of the day, run the savegroup -O to ensure the index is backed-up
and it will go to the pool associated with the group.
you don't have to create another pool with this.
cheers - han
At 12/4/2002 09:43 AM, Robert Loh wrote:
As mentioned I don't want all clients going to the pool that backup only
indexes. May be I should give an example to illustrate myself clearly.
This is in DDS SAN environment with 1 backup server and 2 storage node
sharing only 1 HP LTO drive.
Backup server name = BServer1
Storage Node one name = SN1
Storage Node two name = SN2
If I put all of them together in one group, they will be race condition
depending who own the drive first then only backup it's own data, for
instant Bserver1 own the drive so SN1 and SN2 has to wait till Bserver1
completed backup then unmount and SN1 need to mount the some tape again and
backup continue and so on.....That might cause the whole group in confuse
situation.
So what I did is separate them in different groups as below
Bserver1 in GroupA
SN1 in GroupB
SN2 in GroupC
Then I will hit another problem, After backing up GroupB, SN1 need to umount
tape then Bserver1 hold the drive and mount tape to just only backup indexes
for SN1, that consume a lot of time. That is why I need to configure SN1 and
SN2 to backup to "file" device then archive them to tape once a week.
I have setup a Pool call IndexPool with below parameters
Saveset = index
Client = SN1 and SN2
device=D:\nsr\indexpool
One funny thing is when I issue command C:\savegrp -pvvv GroupB it saying
that the SN1:index will go to indexpool and data will go to OnsitePool, that
is exactly what I want. But when I trigger the group GroupB backup, all just
going to OnsitePool including the SN1: index then Bserver has to mount the
tape to backup SN1:index.
My question is why Networker 6.1.2 pull by client doesn't work and
discrepancy info given by savegrp command.
Thanks
-----Original Message-----
From: Stan Horwitz [mailto:stan AT TEMPLE DOT EDU]
Sent: Tuesday, December 03, 2002 10:13 PM
To: NETWORKER AT LISTMAIL.TEMPLE DOT EDU
Subject: Re: [Networker] FW: [Networker] Pull by Pool doesn't work for
6.1.2
On Tue, 3 Dec 2002, Dirk Verhoeff wrote:
> You could make a group that holds all clients, assign it's own pool, label
> volumes for that pool on a filesystemdevice and then
> run the group from cron with:
>
> savegrp -O groupname
>
> This will only backup the indexes and the bootstrap information.
I use multiple tape pools. I also do a "savegrp -O groupname" twice a
week. How can I stop NSR from complaining when I reconfigure a pool
to not back up its indices?
--
Note: To sign off this list, send a "signoff" 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" 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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
cheers - han
--
Note: To sign off this list, send a "signoff" 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.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
