Hi,
On 2/27/2006 2:32 PM, Baptiste Malguy wrote:
Hello,
Anyone willing to advise me for my specific issue please ? For the
moment, I hold on blocked by this problem, unfortunately, and I would
appreciate so much some help.
Thanks guys.
Baptiste Malguy a écrit :
Hello,
I have a question related to cloning jobs.
I've just re-read my mail before pushing the "Send" button, and ... a
little long I'm afraid.
I want to write my full backups on two different storages (disk + DLT4).
To achieve this goal, I use the Run attribute of the Job resource,
providing the same job name: that's job cloning.
The job is normally an incremental one, that becomes a full one once a
month thanks to the corresponding schedule.
Only the monthly full backups have to go on the tapes. Not the
incremental ones neither the potential full ones that may occur because
FileSet resource has been modified.
And that's the point: how to avoid these random full backups to go on
the tape, that is, how to prevent the original job (writing to disk) to
clone itself (to write on tape).
I've never actually used the python interface, but it might be possible
to do this using python events.
I thought about using Run Before Job attribute, to prevent the cloned
job to work returning a non-zero value, but I don't know what to provide
to the run command to let it know to return 0 or non-zero.
The command needs:
- "I'm a clone" or "I'm writing to tapes"
No idea if this is possible, but you mght look for a running job,
started a short time ago, with the same name in the catalog. The
destination should be easier to find.
- and "I'm a full backup"
Thats already possible with the objects Bacula provides to its events, I
think.
- and "current date" (because of random full backups)
Rather simple in python :-) although having access to the schedules
might be the more thorough solution.
I can't provide to the command, as long as I know, "I'm a clone" neither
"I'm writing to tapes".
This might not be much help, but perhaps a starting point for you.
Arno
Here are some pieces of the setup:
Job {
Name = server-test
Write Bootstrap = /var/lib/bacula/server-test.bsr
JobDefs = default
FileSet = server-test
Client = server-fd
Run = "server-test-etc level=\"%l\" since=\"%s\" pool=cart-month
storage=dlt-storage"
Type = Backup
Storage = file-storage
Messages = standard
Priority = 10
Pool = file-full
}
Schedule {
Name = weekly-cycle
Run = Full Pool=file-full 1st sat at 1:05
Run = Differential Pool=file-diff 2nd-5th sat at 1:05
Run = Incremental Pool=file-inc sun-fri at 1:05
}
JobDefs {
Name = default
Type = Backup
Level = Incremental
FileSet = full
Schedule = weekly-cycle
Storage = file-storage
Messages = standard
Priority = 10
Pool = default
}
I've added this to see what I obtain:
RunBeforeJob = "/etc/bacula/scripts/cloned-inc-not-on-tape \"%c\"
\"%d\" \"%i\" \"%e\" \"%j\" \"%l\" \"%n\" \"%t\" \"%v\""
And the command line is something like:
$1="server-fd"
$2=""
$3="634"
$4="Unknown term code"
$5="server-test-etc.2006-02-22_11.30.31"
$6="Full"
$7="Backup"
$8="Backup"
$9=""
From this point, hoping I've been clear enought (not sure), does anyone
has an idea ?
Bye the way, I use Bacula 1.38.3. I'm looking forward for the project
item #2 "Implement Migration that moves Jobs from one Pool to
another".
Thank you and keep on improving this great software !
--
IT-Service Lehmann [EMAIL PROTECTED]
Arno Lehmann http://www.its-lehmann.de
-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid0944&bid$1720&dat1642
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
