Picon
Favicon

Re: What will i do if a backup is not successfull?

* Alexander Jolk <alexj <at> buf.fr> [20050606 16:17]:
> Jean-Francois Malouin wrote:
> 
> >    su amanda -c "/opt/amanda/amanda6/sbin/amdump archival-conf6 bullcalf \
> >    /data/ipl/ipl10/broche/colindata"
> >
> >The DLE from the disklist:
> >
> >bullcalf /data/ipl/ipl10/broche/colindata /data/ipl/ipl10/broche/colindata 
> >{
> >    tar-archival
> >    exclude append "./ave/*"
> >    exclude append "./brainvisa/*"
> >    exclude append "./pd/*"
> >    exclude append "./t1/*"
> >    exclude append "./t2/*"
> >    exclude append "./vaisseaux/*"
> >}
> >
> >
> >Problem: that DLE has exclude list and it seems that all excluded
> >subdirs are making it to tape. Is this the normal/expected behaviour?
> 
> Someone correct me if I'm wrong, but I believe the DLE 
> /data/ipl/ipl10/broche/colindata is interpreted as matching expression 
> just like `amadmin find' does it.  Which means that every DLE on host 
> bullcalf starting with this string would get dumped, as individual DLEs 
> of course.  Is that what you observe?  In that case, simply adding a 
> dollar sign `$' to the end of the DLE will stop it from matching at the 
> beginning.
(Continue reading)

Jon LaBadie | 7 Jun 02:22

Re: no backups yesterday?

On Mon, Jun 06, 2005 at 04:08:03PM -0600, Cam wrote:
> 
> oops... here's the problem... i did copy the crontab from the docs,
> which only run 2-6... why is that?  Shouldn't amanda run everyday?
> 
> > I am really sorry to sound like some annoyed core-team-member, but
> > right now I am feeling unable to hide my unknowingness of your current
> > setup.

Not if someone wants to run amdump once a week.
Not if someone wants to run amdump on weekdays.
Not if someone wants to run amdump six times a week (like me).

--

-- 
Jon H. LaBadie                  jon <at> jgcomp.com
 JG Computing
 4455 Province Line Road        (609) 252-0159
 Princeton, NJ  08540-4322      (609) 683-7220 (fax)

Picon
Favicon

handling off-site tape storage

Hi,

I'm curious how admins out there handle the off-site storage situation.  The
requirement here is that full weekly backups are stored off-site for 4
weeks, full monthly backups for 6 months etc.

Obviously letting Amanda decide when level 0 should be done on a DLE doesn't
give you a bunch of tapes on Friday with all DLE's level 0.

The previous admin forced the situation with the amadmin force option.  Is
this the best way?  Is there another way?  I'm happy to try and change the
full-friday mindset here if there is a better solution.

Appreciate any insight.

Regards,
Greg.

Frank Smith | 7 Jun 06:43
Favicon

Re: handling off-site tape storage

--On Tuesday, June 07, 2005 13:13:57 +1000 "Keenan, Greg John (Greg)** CTR **" <gjkeenan <at> lucent.com> wrote:

> Hi,
> 
> I'm curious how admins out there handle the off-site storage situation.  The
> requirement here is that full weekly backups are stored off-site for 4
> weeks, full monthly backups for 6 months etc.
> 
> Obviously letting Amanda decide when level 0 should be done on a DLE doesn't
> give you a bunch of tapes on Friday with all DLE's level 0.
> 
> The previous admin forced the situation with the amadmin force option.  Is
> this the best way?  Is there another way?  I'm happy to try and change the
> full-friday mindset here if there is a better solution.
> 

You could run a separate config (using always-full and record no) weekly
and take that set off-site.

Another option would be to take a dumpcycle's worth of tapes off-site; there
should be at least one full of each DLE in the set.

If you have a remote location with a tape drive (and adequate bandwidth), you
can run an Amanda config remotely.  Then the tapes are off-site as soon as
they are written.  For a shorter backup window, do the full backups to an
on-site disk, then backup that remotely.

I'm sure there are other ways as well.

Frank
(Continue reading)

Mike Delaney | 7 Jun 08:04

Re: handling off-site tape storage

On Tue, Jun 07, 2005 at 01:13:57PM +1000, Keenan, Greg John wrote:
> 
> I'm curious how admins out there handle the off-site storage situation.  The
> requirement here is that full weekly backups are stored off-site for 4
> weeks, full monthly backups for 6 months etc.

For the weekly fulls, I'd just send the whole dumpcycle offsite (assuming
dumpcycle=1week) for 4 weeks.  My preference in that situation would be to
have the current week's tapes in the library, the previous week onsite
in a firesafe in case they're needed to restore something recent, and the
remainder offsite.

For the monthlys, I run a separate config with dumpcycle=0 (i.e. always full)
against the same disklist as my regular nightly config.  Those get sent
offsite.  For yearly archives, pick one month out of that set, mark the
tapes noreuse, and keep them offsite for a longer period.

Paul Bijnens | 7 Jun 11:26
Favicon
Gravatar

Re: Fwd: one box won't back up!

Cam wrote:
> 
>>Had the exact same problem last week. Solution was to modify
>>Checkpoint FW-1 Firewall rules to time out UDP packets after 10
>>minutes instead of the default 40 seconds.
>>
>>If you use another stateful firewall, look at it's documentation. The
>>problem lies in the fact that there is a timeout with UDP packets.
> 
> 
> I'm just running a fairly simple iptables firewall... any clues on how
> to change the timeout for UDP packets (i always thought this would
> have been a thing i would tweak in amanda).

Iptables cannot change it for all udp packets (would be a bad idea too),
but the good news is that the ip_conntrack_amanda module can tune it,
specifically for amanda, using the "master_timeout=..." parameter:

http://www.amanda.org/docs/faq.html#id2554525

(The default UDP timeout in iptables is 180 seconds.)

--

-- 
Paul Bijnens, Xplanation                            Tel  +32 16 397.511
Technologielaan 21 bus 2, B-3001 Leuven, BELGIUM    Fax  +32 16 397.512
http://www.xplanation.com/          email:  Paul.Bijnens <at> xplanation.com
***********************************************************************
* I think I've got the hang of it now:  exit, ^D, ^C, ^\, ^Z, ^Q, F6, *
* quit,  ZZ, :q, :q!,  M-Z, ^X^C,  logoff, logout, close, bye,  /bye, *
* stop, end, F3, ~., ^]c, +++ ATH, disconnect, halt,  abort,  hangup, *
(Continue reading)

Cedric BRINER | 7 Jun 14:24
Picon

latops

hi,

Does some of you are backuping laptops with amanda.

If yes.. how do you do it?
otherwise
which tool do you use it (has to work for M$, Linux, Mac)

thank for your time

Ced.
--

-- 

Cedric BRINER

Matt Hyclak | 7 Jun 14:13

Re: latops

On Tue, Jun 07, 2005 at 02:24:43PM +0200, Cedric BRINER enlightened us:
> Does some of you are backuping laptops with amanda.
> 
> If yes.. how do you do it?
> otherwise
> which tool do you use it (has to work for M$, Linux, Mac)
> 

Look into unison. You can sync to a hard disk somewhere which is accesible
to amanda.

Matt

--

-- 
Matt Hyclak
Department of Mathematics 
Department of Social Work
Ohio University
(740) 593-1263
Cam | 7 Jun 17:09
Picon

ERROR unknown service: noop

Hi guys,

Initially i had thought that amanda was failing to backup one of my
clients w/out an error, but i was wrong.  in all of the
/tmp/amandad*000.debug files there is i have this output:

Amanda 2.4 REQ HANDLE 000-C0900608 SEQ 1118089786
SECURITY USER amanda
SERVICE noop
OPTIONS features=fffffeff9ffe0f;
--------

sending nack:
----
Amanda 2.4 NAK HANDLE 000-C0900608 SEQ 1118089786
ERROR unknown service: noop

what is noop?

thanks,
Cameron Matheson

khalid maqsudi | 7 Jun 17:41
Picon

disklist help

Hello,
 
What should my disklist look like if I want to dump the entire c1t0d0s5 slice (which is also my holding disk) to tape.   I have only 1 machine (master) and no clients connected to a powervault 128T tape device running solaris 8.  Any help would be appreciated. Thanks.  Also,  How can I verify if I have the client software on my master (or was that taken care of automatically when I compiled AMANDA on it as a master? )  Thanks.
 
 
-K

Gmane