Re: [Networker] savepnpc that alters the list of filesystems?
2013-02-06 18:04:43
In regard to: RE: [Networker] savepnpc that alters the list of...:
I don't think that will work, the process as far as I know is that
savegrp kicks off a savefs process on the client which probes the
client for valid savesets, these are then passed to individual save
commands (or in your case savepnpc commands), so the file system list is
already defined by the time the savepnpc process is run
That's what we were afraid of.
Thanks to you and Stan for confirming. I went ahead and set up a
dual-client config. I prefer to avoid that when possible, but we've
used it before for other hosts.
Backup group #1 has client definition #1, which lists "/bbsnap" as the
only saveset and has savepnpc as the backup command.
Backup group #2 has client definition #2, which lists "All" as the
saveset, nothing for the backup command, and has a custom directive
that includes all of the "Unix standard directives" plus it adds
<< /bbsnap >>
+skip: .?* *
I know that there have been some arguments in the past about whether
it should be +skip or +null when using a dual or split-client config,
but I have yet to encounter any problems when using +skip with half
of a split-client config.
Thanks again for the information you've both provided!
Tim
bash-3.2# savegrp -I -vvv -c your_client -l incr -G YOUR_GROUP
73605:savegrp: YOUR_GROUP: Savegrp is started with hard limit of 1430 minutes
on maximum duration
90528:savegrp: your_client:All level=incr
7236:savegrp: Group will not limit job parallelism
83643:savegrp: your_client:All started
savefs -s backup_server-c your_client -g YOUR_GROUP -p -l full -R -v
your_client:C:\ level=incr, vers=pools, p=2
your_client:VSS SYSTEM BOOT:\ level=incr, vers=pools, p=2
your_client:VSS SYSTEM SERVICES:\ level=incr, vers=pools, p=2
your_client:VSS USER DATA:\ level=incr, vers=pools, p=2
your_client:VSS OTHER:\ level=incr, vers=pools, p=2
your_client:VSS SYSTEM FILESET:\ level=incr, vers=pools, p=2
your_client:VSS ASR DISK:\ level=incr, vers=pools, p=2
90491:savegrp: your_client:All succeeded.
83647:savegrp: YOUR_GROUP your_client:All See the file
/nsr/logs/sg/YOUR_GROUP/4279 for command output
83643:savegrp: your_client:VSS ASR DISK:\ started
save -s backup_server-g YOUR_GROUP -LL -m your_client -t 1360141321 -l incr -W 78 -N "VSS ASR
DISK:\\" "VSS ASR DISK:\\"
83643:savegrp: your_client:VSS SYSTEM FILESET:\ started
save -s backup_server-g YOUR_GROUP -LL -m your_client -t 1359709582 -l incr -W 78 -N "VSS
SYSTEM FILESET:\\" "VSS SYSTEM FILESET:\\"
83643:savegrp: your_client:VSS OTHER:\ started
save -s backup_server-g YOUR_GROUP -LL -m your_client -t 1360141370 -l incr -W 78 -N "VSS
OTHER:\\" "VSS OTHER:\\"
83643:savegrp: your_client:VSS USER DATA:\ started
save -s backup_server-g YOUR_GROUP -LL -m your_client -t 1360141379 -l incr -W 78 -N "VSS USER
DATA:\\" "VSS USER DATA:\\"
83643:savegrp: your_client:VSS SYSTEM SERVICES:\ started
save -s backup_server-g YOUR_GROUP -LL -m your_client -t 1360141425 -l incr -W 78 -N "VSS
SYSTEM SERVICES:\\" "VSS SYSTEM SERVICES:\\"
83643:savegrp: your_client:VSS SYSTEM BOOT:\ started
save -s backup_server-g YOUR_GROUP -LL -m your_client -t 1360141428 -l incr -W 78 -N "VSS
SYSTEM BOOT:\\" "VSS SYSTEM BOOT:\\"
83643:savegrp: your_client:C:\ started
save -s backup_server-g YOUR_GROUP -LL -m your_client -t 1360141494 -l incr -W 78 -N
"C:\\" "C:\\"
It might work if you create a separate client definition (in a separate group) that
specifies the file system that you want backed up, and use the savepnpc process to
create the file system ... the example below fails because the "test"
filesystem / file don't exist, but it does mean that it is passed to the save (or
savepnpc) process.
bash-3.2# savegrp -I -vvv -c your_client -l incr -G DD_YOUR_GROUP
73605:savegrp: DD_YOUR_GROUP: Savegrp is started with hard limit of 1430
minutes on maximum duration
90528:savegrp: your_client:test level=incr
7236:savegrp: Group will not limit job parallelism
83643:savegrp: your_client:savefs started
savefs -s backup_server -c your_client -g DD_YOUR_GROUP -p -l full -R -v -F test
your_client:test level=full, vers=pools, p=2
90491:savegrp: your_client:savefs succeeded.
83647:savegrp: DD_YOUR_GROUP your_client:savefs See the file
/nsr/logs/sg/DD_YOUR_GROUP/4297 for command output
83643:savegrp: your_client:test started
save -s backup_server -g DD_YOUR_GROUP -LL -m your_client -l full -W 78 -N test
test
84078:savegrp: command 'save -s backup_server -g DD_YOUR_GROUP -LL -m
your_client -l full -W 78 -N test test' for client your_client exited with
return code 1
77562:savegrp: job (4298) host: your_client savepoint: test had ERROR
indication(s) at completion
90491:savegrp: your_client:test failed.
Mat
-----Original Message-----
From: EMC NetWorker discussion [mailto:NETWORKER AT LISTSERV.TEMPLE DOT EDU] On
Behalf Of Tim Mooney
Sent: Thursday, 7 February 2013 6:58 AM
To: NETWORKER AT LISTSERV.TEMPLE DOT EDU
Subject: [Networker] savepnpc that alters the list of filesystems?
All-
NetWorker server: NetWorker 7.6.2.5 on RHEL 6.3 x86_64 NetWorker client:
NetWorker 7.6.2.5 on RHEL 5.9 x86_64
Does anyone know when, during the save process, the list of filesystems to back
up is set? In particular, does it happen *before* the precmd in a savepnpc
script?
I ask because we're trying to use a savepnpc precmd script to snapshot a Linux LVM volume
and then mount the snapshot in a new place and have NetWorker back up the snapshot. We
are using the "All" saveset for the client. So far, NetWorker hasn't even
noticed that the new snapshot filesystem is available for backup, so I'm concerned that
it may be setting the list of volumes to back up *before* the precmd runs.
Note that the precmd is doing the snapshot, mounting the volume, and updating
/etc/fstab with the entry for the new mounted volume.
Any thoughts on whether this should work?
|
|
|