Amanda-Users

Re: taper write error even though amcheck -sclt says ok.

2005-03-10 23:14:33
Subject: Re: taper write error even though amcheck -sclt says ok.
From: Jon LaBadie <jon AT jgcomp DOT com>
To: amanda-users AT amanda DOT org
Date: Thu, 10 Mar 2005 23:09:24 -0500
On Fri, Mar 11, 2005 at 01:48:05AM +0000, Bruce S. Skinner wrote:
> Jon LaBadie <jon AT jgcomp DOT com> writes:
> 
> > On Thu, Mar 10, 2005 at 03:51:49PM +0000, Bruce S. Skinner wrote:
> > > Hi,
> > > 
> > > /var/log/amanda/Norstead/log:
> > > 
> > >   DISK planner alsike alsike_root
> > >   DISK planner alsike alsike_boot
> > >   DISK planner alsike alsike_archive_a
> > >   DISK planner alsike alsike_archive_b
> > >   DISK planner alsike alsike_home
> > >   START planner date 20050310
> > >   WARNING planner tapecycle (8) <= runspercycle (28)
> > 
> > That doesn't sound good. Says you have 8 tapes and earlier you
> > said you might use 3 per run.  Further, this says you might
> > run dumps 28 times between level 0 dumps.  Not good at all.
> 
> I know it looks bad on the surface, however I know I can dump this
> whole system up on 3 tapes and then the daily dumps won't be all
> that large.   So my plan is after seeing that the full dumps make it
> to tape, I'll only  stick a tape in the drive every week letting the
> incremental dumps build up on /var/spool/amanda.  Even though
> tapecycle is 8 their are 12 tapes on the shelf.  Does this make sense,
> or am I out to lunch?


You are aware, correct, that amanda will NOT schedule full dumps on day 1
and incrementals for the next 27 days?  While each DLE will get a full
dump the first time it is backed up, after that (starting on day 2) it
will schedule for each day a mix of some DLEs incremental and others full.


> > >   taperalgo largestfit  # Do The largest dump image that will fit on the
> > >                         # current tape next.
> > >   
> > >   diskdir "/var/spool/amanda"  # where the holding disk is
> > 
> > 
> > What is this, another keyword I did not see sneak in?
> 
> You mean diskdir?  The path to the holding disk?


Never saw it used before.  A grep did find it described in a historical document
misleadingly titled "WHAT'S NEW".  However, none of the current example files
or man pages mention it.  The way to define each holding disk with which I'm
familiar is via stanzas such as:

    holdingdisk hd2 {
        comment "second holding disk on big IDE drive"
        directory "/w3/dumps/amanda/DS1"
        use -2 Gb
        # chunksize 1 Gb
    }

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