Re: mixed full / incremental backup...
Hi Stefan et al,... ...and at first, thanks for all the input - much appreciated! :) On Tue, 9 Mar 2004 11:17:26 +0100 "Stefan G. Weichinger" <[EMAIL PROTECTED]> wrote: > Wrong syntax there ... > > Doesn't amcheck complain ?? Not really, everything seems to be fine, though I made some modifications in my configuration according to the hints I gathered here. Anyhow, in the end I guess it all comes down to some problems with our autoloader working with chg-scsi, sometimes ending amcheck / amdump without finding the correct tape, sometimes (like this morning) freezing "a chg-scsi -slot current" process for hours (backup scheduled to start at 6:00 am, still waiting in right this situation at 9:30 am ... :/ ). I'll play around with the changer debugging options and see what happens... :( Cheers, Kris -- Kristian Rink -- Programmierung/Systembetreuung planConnect GmbH * Strehlener Str. 12 - 14 * 01069 Dresden 0176 24472771 * [EMAIL PROTECTED]
Re: mixed full / incremental backup...
Here's a very simple solution: 1. "reserve 100" in amanda.conf (or comment out "reserve" line) 2. leave out the tape during the week (*or change dev to /no/such/tape) 3. run "amflush" before Friday backups In this case, amanda should try to do "degraded mode" backups during the week while there is no tape, only incrementals. On Friday, you put a tape in the drive before the backups, and I think it should try to catch up on some full backups. It probably won't give you perfect incr/full separation, but it would be good enough for me. Naturally, you will need sufficient holding disk space. You will have to keep an eye out for messages warning about "overwrite last full dump." * do not change tapedev to /dev/null because that activates a "test mode" where amanda will happily dump all data into the bit bucket rather than decide there is no tape. --jonathan Jon LaBadie wrote: What about a single config. Set it up with a long dumpcycle. Use dumptypes that specify "always incremental", or "no-full", or "strategy incr" or ??? I.e. something that reduces the possibility of full dumps.
Re: mixed full / incremental backup...
On Tue, Mar 09, 2004 at 09:41:56AM +0100, Kristian Rink wrote: > > Hi all,... > > Since people around here still are stuck wanting to have a > "friday-full/rest-of-the-week-incremental" backup strategy, I read > through > ... > created two amanda configurations with the same disklist for full > and for incremental backup, labeled my tapes and tested the whole > mess. About the full backup, things work pretty fine. However, > trying to run the incremental setup always leaves me with an error > mail such as below: > Just a possibility for an alternative arrangement. I'd hate having to maintain two nearly identical configs for a single disklist. What about a single config. Set it up with a long dumpcycle. Use dumptypes that specify "always incremental", or "no-full", or "strategy incr" or ??? I.e. something that reduces the possibility of full dumps. Then set things up to "force" full dumps on Fridays. Use amadmin commands for this. The force could be done with a separate cron job that runs Fridays before the amdump job or by replacing the amdump command with a shell script that says "if it is friday, first do an amadmin force". A fancier script could check if is anyother day, was a full dump done more recently than last Thursday, if not do the force. This could handle the situation where the friday dumps were not done but you shouldn't wait for two weeks for the next full dumps. -- Jon H. LaBadie [EMAIL PROTECTED] JG Computing 4455 Province Line Road(609) 252-0159 Princeton, NJ 08540-4322 (609) 683-7220 (fax)
Re: mixed full / incremental backup...
Hi, Kristian,
on Dienstag, 09. März 2004 at 10:57 you wrote to amanda-users:
>> You have in your disklist:
>>
>> > localhost work/04 /backup/PlanC3/bf381/docs/work {
>>
>> I am not sure but I would not use something like "work/04" as
>> diskname.
KR> The diskfile is the same for the incremental and the full backup,
KR> and running the full backup works as expected with the same dle's...
KR> At least I'll modify the hostname there to have this fixed.
Ok ...
>> Show us your AMANDA email report ...
KR> (see attached...)
You don't use a holding-disk for root-tar, so it has to fail when
there is no tape ...
I would edit this:
> define dumptype root-tar {
> program "GNUTAR"
> comment "Root partitions dumped with tar"
> options no-compress, index, no-full
> priority high
> }
to this:
define dumptype root-tar {
program "GNUTAR"
comment "Root partitions dumped with tar"
compress none
index yes
strategy nofull
priority high
}
Wrong syntax there ...
Doesn't amcheck complain ??
--
best regards,
Stefan
Stefan G. Weichinger
mailto:[EMAIL PROTECTED]
Re: mixed full / incremental backup...
Hi again,...
On Tue, 9 Mar 2004 10:52:02 +0100
"Stefan G. Weichinger" <[EMAIL PROTECTED]> wrote:
> KR> Okay, thanks for the hint. Anyhow, I am just running amanda to
> KR> back up files from this one host; is the naming likely to get
> KR> me into trouble even this way?
>
> amrecover will not work as expected.
Hmm, this is not a good thing, so I will change this.
[disklist]
> You have in your disklist:
>
> > localhost work/04 /backup/PlanC3/bf381/docs/work {
>
> I am not sure but I would not use something like "work/04" as
> diskname.
The diskfile is the same for the incremental and the full backup,
and running the full backup works as expected with the same dle's...
At least I'll modify the hostname there to have this fixed.
> Show us your AMANDA email report ...
(see attached...)
Cheers,
Kris
--
Kristian Rink -- Programmierung/Systembetreuung
planConnect GmbH * Strehlener Str. 12 - 14 * 01069 Dresden
0176 24472771 * [EMAIL PROTECTED]
planConnect_AMANDA_MAIL_REPORT_FOR_March_9,_2004
Description: Binary data
Re: mixed full / incremental backup...
Hi, Kristian,
on Dienstag, 09. März 2004 at 10:24 you wrote to amanda-users:
>> At first let me tell you that you should not use "localhost" in
>> your DLEs. Use the fully qualified hostname (FQDN) instead.
>> localhost bites you at recovery time.
KR> Okay, thanks for the hint. Anyhow, I am just running amanda to back
KR> up files from this one host; is the naming likely to get me into
KR> trouble even this way?
amrecover will not work as expected.
>> 2. The message does not say that localhost:work/04 does not fit on
>> tape. It says that the size of the whole dump (~ all DLEs) is
>> larger than your tape.
KR> Hmmm, okay, but even this seems to be strange while the full dump
KR> doesn't complain about this and just moves anything found in the
KR> file area to tape...
>> And the disklist?
KR> see attachment. :) Basically, I split up the base directory of our
KR> document management system into several smaller chunks because it
KR> won't take long for the tree to not fit on a single tape anymore...
You have in your disklist:
> localhost work/04 /backup/PlanC3/bf381/docs/work {
I am not sure but I would not use something like "work/04" as diskname.
Edit to something like:
my.host.my.domain work04 /backup/PlanC3/bf381/docs/work {
(get rid of the "/" in "work/04").
--
Does any DLE run through here?
Show us your AMANDA email report ...
--
best regards,
Stefan
Stefan G. Weichinger
mailto:[EMAIL PROTECTED]
Re: mixed full / incremental backup...
Hi Stefan, ...and at first, thanks for your reply. On Tue, 9 Mar 2004 10:02:59 +0100 "Stefan G. Weichinger" <[EMAIL PROTECTED]> wrote: > KR> This is strange - tape Back06 actually _is_ inside the changer > KR> and labeled correctly. > > Does amcheck run through fine ? > Is Back06 in your tapelist? I got that fixed in the meantime finding out that the tape was unloaded to a slot this amanda-setup is not intended to use. Entirely my fault... > KR> doubt that because the full dump itself fits on tape rather > KR> well. > > At first let me tell you that you should not use "localhost" in > your DLEs. Use the fully qualified hostname (FQDN) instead. > localhost bites you at recovery time. Okay, thanks for the hint. Anyhow, I am just running amanda to back up files from this one host; is the naming likely to get me into trouble even this way? > 2. The message does not say that localhost:work/04 does not fit on > tape. It says that the size of the whole dump (~ all DLEs) is > larger than your tape. Hmmm, okay, but even this seems to be strange while the full dump doesn't complain about this and just moves anything found in the file area to tape... > KR> Can anyone point me where to put my eyes to get this fixed? > KR> amanda.conf of the incremental setup is included. > > And the disklist? see attachment. :) Basically, I split up the base directory of our document management system into several smaller chunks because it won't take long for the tree to not fit on a single tape anymore... Thanks / bye, Kris -- Kristian Rink -- Programmierung/Systembetreuung planConnect GmbH * Strehlener Str. 12 - 14 * 01069 Dresden 0176 24472771 * [EMAIL PROTECTED] disklist Description: Binary data
Re: mixed full / incremental backup...
Hi, Kristian, on Dienstag, 09. März 2004 at 09:41 you wrote to amanda-users: >> *** A TAPE ERROR OCCURRED: [label Back06 or new tape not found in >> rack]. Some dumps may have been left in the holding disk. KR> This is strange - tape Back06 actually _is_ inside the changer and KR> labeled correctly. Does amcheck run through fine ? Is Back06 in your tapelist? >> FAILURE AND STRANGE DUMP SUMMARY: >> larger than tape, -1 KB, skipping incremental] localhost >> work/04 lev 1 FAILED [dump larger than tape, -1 KB, skipping >> incremental] localhost project lev 1 FAILED [dump larger than >> tape] KR> This is what always uses to show up in those situations - "dump KR> larger than tape" for all the dle's I used to set up. I doubt that KR> because the full dump itself fits on tape rather well. At first let me tell you that you should not use "localhost" in your DLEs. Use the fully qualified hostname (FQDN) instead. localhost bites you at recovery time. 2. The message does not say that localhost:work/04 does not fit on tape. It says that the size of the whole dump (~ all DLEs) is larger than your tape. KR> Can anyone point me where to put my eyes to get this fixed? KR> amanda.conf of the incremental setup is included. And the disklist? -- best regards, Stefan Stefan G. Weichinger mailto:[EMAIL PROTECTED]
mixed full / incremental backup...
Hi all,... Since people around here still are stuck wanting to have a "friday-full/rest-of-the-week-incremental" backup strategy, I read through http://www.backupcentral.com/amanda-19.html created two amanda configurations with the same disklist for full and for incremental backup, labeled my tapes and tested the whole mess. About the full backup, things work pretty fine. However, trying to run the incremental setup always leaves me with an error mail such as below: > *** A TAPE ERROR OCCURRED: [label Back06 or new tape not found in > rack]. Some dumps may have been left in the holding disk. This is strange - tape Back06 actually _is_ inside the changer and labeled correctly. > > FAILURE AND STRANGE DUMP SUMMARY: > larger than tape, -1 KB, skipping incremental] localhost > work/04 lev 1 FAILED [dump larger than tape, -1 KB, skipping > incremental] localhost project lev 1 FAILED [dump larger than > tape] This is what always uses to show up in those situations - "dump larger than tape" for all the dle's I used to set up. I doubt that because the full dump itself fits on tape rather well. Can anyone point me where to put my eyes to get this fixed? amanda.conf of the incremental setup is included. TIA, have a fine day everyone. Kris -- Kristian Rink -- Programmierung/Systembetreuung planConnect GmbH * Strehlener Str. 12 - 14 * 01069 Dresden 0176 24472771 * [EMAIL PROTECTED] amanda.conf Description: Binary data
