Amanda-Users

Re: dumpcycle, dumplevel

2003-07-31 10:25:42
Subject: Re: dumpcycle, dumplevel
From: Brian Cuttler <brian AT wadsworth DOT org>
To: gene.heskett AT verizon DOT net
Date: Thu, 31 Jul 2003 10:23:02 -0400 (EDT)
Hello Gene,

Yes, we have moved away from root user to "amanda" user. This was an
issue for earlier versions of amanda on earlier versions of IRIX (but
we used bin on Sun). Newer installations use amanda for both architectures.

Runspercyle - can't imagine why that isn't in the file, I don't think
its the whole of the problem though as this particular instance of 
amanda has been running for quite some time.

I forced a level 0 on the problem partition and started amdump.
Finding is that the level 0 is estimated to exceed the capacity
of the tape drive.

I suppose the proper solution is to install the latest amanda
2.4.4-(what?), make sure I include runspercycle and modify the
DLE for /usr2 (and /usr3) to specify TAR DLE for segments of
these two partitions. Converting to user amanda as I go.

Hadn't realized that that partition had grown so (wishing the 
users where better about storage... but I seem to maintain a 
lot of unrealistic views).

SPAN-tapes...

I missed the warnings that the level 0 was overdue.

My jukebox could really use a multi-tape option that would split
usage of the drives (one/jukebox) so that DLEs where written to 
first available drive ie: two tapers balancing DLE across the 
tapes -- not a raid usage, just maximizing throughput.

Not that it matters here, this is a simple capacity issue on a
non-jukebox system. The only system I have has two jukeboxes but
no way to maximize thoughput from a single amanda instance.

                                                thanks,

                                                Brian

> On Thursday 31 July 2003 09:33, Brian Cuttler wrote:
> >Hello Amanda users,
> >
> >Well, this is a first (for our site anyway).
> >
> From amanda.conf
> >
> >dumpuser "root"         # the user to run dumps under
> 
> No, you cannot use 'root' above.  You must use an unpriviledged user 
> such as 'amanda' who is made a member of a high level group like 
> disk.  Please read the docs.
> 
> >inparallel 8            # maximum dumpers that will run in parallel
> >netusage  800           # maximum net bandwidth for Amanda, in KB
> > per sec
> >
> >dumpcycle 1 weeks       # the number of days in the normal dump
> > cycle
> >tapecycle 25 tapes      # the number of tapes in rotation
> >
> >bumpsize 10 MB          # minimum savings (threshold) to bump level
> > 1 -> 2 bumpdays     2          # minimum days at each level
> >bumpmult     2          # threshold = bumpsize * (level-1)**bumpmult
> >
> >Ok, maybe bumpsize is a little small... still, if its a dumpcycle of
> >one week and we have bumpdays of 2 then I'd never expect to see
> >anything above a level 3 dump.
> >
> >I'd thought there was a parameter for dumps/dumpcycle, in our case 5
> >dumps during the weekly dump cycle.
> 
> Thats "runspercycle 5" in your case. Its missing above.
> 
> >Ok, maybe I'm not understanding how bumpmult is used.
> >
> >I'm surprised that the dump level isn't reset, ie:
> >     dumplevel = min(calculated_level, 9)
> >
> >and that planner returned nothing (empty) for all other partitions.
> >
> >----- Forwarded message from Super-User -----
> >
> >FAILURE AND STRANGE DUMP SUMMARY:
> >  planner: FATAL error [planner askfor: lev out of range -1..10: 10]
> >  laksha     /dev/root RESULTS MISSING
> >  laksha     /usr1 RESULTS MISSING
> >  laksha     /usr2 RESULTS MISSING
> >  laksha     /usr3 RESULTS MISSING
> >  andaman    /dev/root RESULTS MISSING
> >
> >NOTES:
> >  driver: WARNING: /amanda/work: 92160000 KB requested, but only
> > 87586720 KB available. driver: WARNING: got empty schedule from
> > planner
> >  taper: tape LAKSHA13 kb 0 fm 0 [OK]
> >
> >DUMP SUMMARY:
> >                                      DUMPER STATS               
> > TAPER STATS HOSTNAME DISK           L   ORIG-KB    OUT-KB COMP%
> > MMM:SS   KB/s MMM:SS   KB/s ------------------------
> > --------------------------------------- ------------- andaman
> > /dev/root        MISSING
> > --------------------------------------------- laksha  /dev/root    
> >    MISSING --------------------------------------------- laksha 
> > /usr1            MISSING
> > --------------------------------------------- laksha  /usr2        
> >    MISSING --------------------------------------------- laksha 
> > /usr3            MISSING
> > ---------------------------------------------
> >
> >(brought to you by Amanda version 2.4.2p2)
> 
> And this version, while serviceable, is now VERY long in the tooth.
> 
> >----- End of forwarded message from Super-User -----
> >
> >yah, I'm way missing my level 0 for /usr2
> >
> >laksha 13# /usr/local/sbin/amadmin laksha find laksha /usr2
> >Scanning /amanda/work...
> >
> >date       host   disk  lv tape or file file status
> >2003-06-26 laksha /usr2  6 LAKSHA14        3 OK
> >2003-06-27 laksha /usr2  6 LAKSHA15        3 OK
> >2003-06-30 laksha /usr2  6 LAKSHA16        2 OK
> >2003-07-01 laksha /usr2  6 LAKSHA17        3 OK
> >2003-07-02 laksha /usr2  6 LAKSHA18        2 OK
> >2003-07-03 laksha /usr2  6 LAKSHA19        3 OK
> >2003-07-04 laksha /usr2  6 LAKSHA20        4 OK
> >2003-07-07 laksha /usr2  6 LAKSHA21        2 OK
> >2003-07-08 laksha /usr2  6 LAKSHA22        3 OK
> >2003-07-09 laksha /usr2  6 LAKSHA23        2 OK
> >2003-07-10 laksha /usr2  6 LAKSHA24        3 OK
> >2003-07-11 laksha /usr2  6 LAKSHA25        3 OK
> >2003-07-14 laksha /usr2  6 LAKSHA01        1 OK
> >2003-07-15 laksha /usr2  6 LAKSHA02        3 OK
> >2003-07-16 laksha /usr2  6 LAKSHA03        2 OK
> >2003-07-17 laksha /usr2  6 LAKSHA04        3 OK
> >2003-07-18 laksha /usr2  6 LAKSHA05        3 OK
> >2003-07-21 laksha /usr2  6 LAKSHA06        2 OK
> >2003-07-22 laksha /usr2  7 LAKSHA07        3 OK
> >2003-07-23 laksha /usr2  7 LAKSHA08        2 OK
> >2003-07-24 laksha /usr2  8 LAKSHA09        3 OK
> >2003-07-25 laksha /usr2  8 LAKSHA10        4 OK
> >2003-07-28 laksha /usr2  9 LAKSHA11        4 OK
> >2003-07-29 laksha /usr2  9 LAKSHA12        3 OK
> >
> >
> >                                             thanks,
> >
> >                                             Brian
> >
> >---
> >   Brian R Cuttler                 brian.cuttler AT wadsworth DOT org
> >   Computer Systems Support        (v) 518 486-1697
> >   Wadsworth Center                (f) 518 473-6384
> >   NYS Department of Health        Help Desk 518 473-0773
> 
> To cap this, the tarball must be unpacked by the user, not root, and 
> built by the user too.  root is only used to do the install.
> 
> Because its an easy mistake to make, I plagerized someones little 
> helper and added this to my personal 'gh.cf' script that configures 
> each new build for me:
> -----------------------------------
> #!/bin/sh
> # since I'm always forgetting to su amanda...
> if [ `whoami` != 'amanda' ]; then
>       echo
>       echo "!!!!!!!!!!!! Warning !!!!!!!!!!!!"
>       echo "Amanda needs to be configured and built by the user amanda,"
>       echo "but must be installed by user root."
>       echo
>       exit 1
> fi
> make clean
> rm -f config.status config.cache
> ./configure --with-user=amanda \
>       --with-group=disk \
>       --with-owner=amanda \
>       --with-tape-device=/dev/nst0 \
>       --with-changer-device=/dev/sg1 \
>       --with-gnu-ld --prefix=/usr/local \
>       --with-debugging=/tmp/amanda-dbg/ \
>       --with-tape-server=coyote.coyote.den \
>       --with-amandahosts \
>       --with-configdir=/usr/local/etc/amanda
> 
> make
> ------------------------------------
> Once this script has run as user amanda, then root does a "make 
> install".  It all takes about 3:30 on this machine.  Modify it to 
> suit your situation of course.
> 
> -- 
> Cheers, Gene
> AMD K6-III@500mhz 320M
> Athlon1600XP@1400mhz  512M
> 99.27% setiathome rank, not too shabby for a WV hillbilly
> Yahoo.com attornies please note, additions to this message
> by Gene Heskett are:
> Copyright 2003 by Maurice Eugene Heskett, all rights reserved.
> 


<Prev in Thread] Current Thread [Next in Thread>