On Wednesday 06 April 2005 17:14, Vicki Stanfield wrote:
>Gene Heskett wrote:
>>On Wednesday 06 April 2005 09:34, Vicki Stanfield wrote:
>>>I am still trying to understand and improve my inherited amanda
>>>configuration. I notice that in the disklist, there are parent
>>>directories listed before their children. Here is an example:
>>>
>>>myhost / comp-low-tar
>>>myhost /usr comp-low-tar
>>>myhost /var comp-med-tar
>>>myhost /net/myhost/home comp-med-tar
>>
>>Hoo boy! Vicky, do you not understand that / and /usr are very
>> likely two entirely seperate filesystems, mounted individually?
>>
>>>Does Amanda backup the subdirectories of an item listed in the
>>>disklist?
>>
>>Unless you explicitely exclude a subdir, it will be recursively
>> backed up by tar. dump OTOH, can only do the full filesystem.
>> Thats why we like tar, much better control.
>>
>>>And does that mean that I am backing up /usr twice? I
>>>suspect that the answer is no, but I would like to confirm this.
>>
>>Look at the /etc/fstab on that machine Vicky, and if there are
>>seperate entries for each, then they are indeed seperate
>> filesystems, and the disklist is correct.
>>
>>>I also notice that some things in the disklist sometimes show up
>>> in amadmin balance as being extraordinarily overdue (like 131
>>> days). due-date #fs orig KB out KB balance
>>>----------------------------------------------
>>>4/06 Wed 5 56836750 46326642 +384.2%
>>>4/07 Thu 2 13680830 13680830 +43.0%
>>>4/08 Fri 0 0 0 ---
>>>4/09 Sat 3 7928350 6604199 -31.0%
>>>4/10 Sun 3 9011080 9011080 -5.8%
>>>4/11 Mon 5 14117700 8890351 -7.1%
>>>4/12 Tue 6 11165980 10039560 +4.9%
>>>4/13 Wed 6 11057670 9254644 -3.3%
>>>
>>>4/16 Sat 8 9326350 8127764 -15.1%
>>>4/17 Sun 4 11941010 10722435 +12.1%
>>>4/18 Mon 1 14422360 14422360 +50.7%
>>>4/19 Tue 3 14011020 11719132 +22.5%
>>>4/20 Wed 6 11232850 8034076 -16.0%
>>>
>>>4/23 Sat 32 10841180 10841180 +13.3%
>>>4/24 Sun 1 14116820 14116820 +47.5%
>>>----------------------------------------------
>>>TOTAL 85 209689950 181791073 9567951
>>> (estimated 19 runs per dumpcycle)
>>>(3 filesystems overdue, the most being overdue 131 days)
>>
>>Thats a bonafide notice that one or more of the filesystems has
>> grown so big that a level 0 on it will no longer fit on the
>> storage medium. When that happens, the backups on that filesystem
>> cease.
>>
>>The former administrators account should contain emails notifying
>> the admin of that problem. I suspect they haven't been read in
>> quite some time. As root, change the ownership from the previous
>> account holder to your's, and then mv the whole thing to your home
>> directory so you can peruse them for clues.
>>
>>So now, your job is to find out which partition has outgrown the
>>media, and convert that disklist entry into a list of subdirs that
>>are each within the capacity of the media to do a level 0 on.
>>
>>I would, at this time, edit only the amanda.conf file and change
>> only the mailto: data so you get these emails. You probably want
>> to inspect the crontab entry for the user who actually executes
>> amanda, and change the mailto: in that, just to make sure you get
>> all the emails it generates and don't have to go charging off into
>> the wilderness without a suitable map of what it is you need to
>> do.
>>
>>>I am suspicious that it is the parent directories that are so much
>>>overdue, but I don't know how to confirm what is overdue.
>>>Additionally, I am looking for a way to determine which disklist
>>>item is going to be fully backed up when. They don't seem to
>>>correspond that way. Is there an option to amadmin that I missed?
>>
>>Probably, but I haven't had occasion to use it for that, so I'm
>> "not a guru(TM)" on amadmins other features.
>>
>>When you do find the perp, then comment that line of the disklist
>> out, and make entries in it for each subdir, doing about half of
>> them for tonight, and the other half tomorrow.
>>
>>As a philosophical thing, I like to keep a disklist entry below 1/2
>> a tape, and many of mine are much smaller than that. Amanda will
>> slowly juggle the schedule until amanda is useing about the same
>> amount of tape for every run. But when the disklist entries are
>> all half a tape or more in size, then amanda has relatively little
>> wiggle room to do those adjustments. Smaller disklist entries
>> result in the ability to do a much finer granularity of adjustment
>> toward that same amount of tape every night goal. It will take
>> amanda 2 or more runs thru the tapecycles worth of tapes before
>> this schedule settles down to only a minor tweek from time to
>> time. You'll get advised of these actions after the fact in the
>> email report mailed to you when amanda is done.
>>
>>>Vicki
>>
>>This seems like its your first encounter with a *nix like system,
>> so things are a wee bit different from the way a winderz box might
>> do it. But you are not rebooting it 4x a day either in case you
>> haven't noted that yet. No BSOD's allowed in this society. :)
>>
>>Good luck.
>
>Not at all. I am just not sure how Amanda behaves. I used a bad
> example. The one I should have used was different in that there
> were exclude files in the subdirectories that were mentioned. Also
> there is a lot of ldap stuff that I'd not encountered before.
>
>Vicki
Sorry, I was making assumptions that obviously aren't true, my
apologies.
As for amanda and ldap stuff, the acronym is known to me and thats
about all. But I'm fairly sure there are those on this list who can
advise.
Have you managed to find and read the mail from amanda yet?
--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.34% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2005 by Maurice Eugene Heskett, all rights reserved.
|
