Jon,

Hard to believe I've run this config as long as I have without
seeing (ie realizing/understanding) the issue.

"reserve" was in fact commented out and it probably the issue here.

[aside]
I have seen the practice in other utilities where commented out values
are listed with their default values, in this case 100 % is the effetive
value of the commented out value, where the comment itself shows a
"suggestion" of 30 %. -- though I admit its an older amanda version and
probably a carry forward config file.

Actually, until recently I ran 2 amanda runs in parallel on the same
box (multiple tape drives) so I was probably masking the issue because
I had an enforced concurrency that I've lost (since one drive failed,
I should say the robot failed) the 2nd amanda config.

I've corrected the reserve value.
We'll see that the next amanda run brings.

                                                thank you,

                                                Brian

On Fri, Jun 23, 2006 at 01:59:27PM -0400, Jon LaBadie wrote:
> On Fri, Jun 23, 2006 at 01:18:15PM -0400, Brian Cuttler wrote:
> > 
> > Hello amanda users,
> > 
> > I'm confused as to why these dumps, where are late in to be
> > dumped today (there are 50 some entries in the disklist) are
> > not making use of the work area.
> > 
> > The current partition is only provide 10 Gig and the work area
> > is 67 Gig and is completely free.
> > 
> > Once earlier/completed DLE finish with the work area the work
> > area is re-used by DLE that are dumped later in the amdump run.
> > Or so I'd thought...
> > 
> > # df -kh /amanda/work
> > Filesystem             size   used  avail capacity  Mounted on
> > /dev/dsk/c7t6d0s0       67G   112M    67G     1%    /amanda
> > 
> > 
> > # /usr/local/sbin/amstatus ninfo | grep dump
> > Using /usr/local/etc/amanda/ninfo/DailySet1/amdump from Thu Jun 22 18:30:00 
> > EDT 2006
> > lyra:/db1             1 10311197k dumping to tape (13:02:00)
> 
> As you are dumping direct to tape.
> 
> My first thought is you have defined your holding disk incorrectly
> (eg syntax or spelling error) or you have "reserve"d too much of it
> for incrementals (the default is 100%) in degraded mode.
> 
> It is also possible you have "holdingdisk no" in your dumptype.
> 
> > lyra:/db2             1 18694369k wait for dumping 
> > lyra:/db3             0 23086605k wait for dumping 
> > lyra:/space           0 31562335k wait for dumping 
> > nnewton:/alm2         0 21446182k wait for dumping 
> > nnewton:/logs-archive 0 70564962k wait for dumping 
> 
> If not software compressed, this DLE will not fit the holding disk
> and will have to go straight to tape.
> 
> > nnewton:/nfsamba2     0 31602082k wait for dumping 
> > nnewton:/nfsamba3     1  5410642k wait for dumping 
> > ranokao:/             0  8188064k wait for dumping 
> > wait for dumping:  15            210555241k           ( 62.25%)
> > dumping to tape :   1             10311197k           (  3.05%)
> > dumping         :   0         0k         0k (  0.00%) (  0.00%)
> > dumped          :  48 127720762k 127684622k (100.03%) ( 37.76%)
> > 7 dumpers idle  : not-idle
> >  dumper0 busy   : 18:17:45  ( 99.98%)
> >  0 dumpers busy :  0:00:00  (  0.00%)
> >  1 dumper busy  : 18:18:00  (100.00%)            not-idle: 18:18:00  
> > (100.00%)
> > 
> 
> -- 
> 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)
---
   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