Hi

I'm trying to set up a backup system for 7 networked machines. Each
machine has an average capacity of about 200GB (not all capacity used
on machines), but due to resource limitations, we're backing up onto
about 480GB of space on the hard drives of our backup server.

I've split the 480GB into 4 virtual tapes of 115GB each (so it
definitely is less than the actual space we have):

------- in my amanda.conf: ---------------------
define tapetype HARD-DISK {
        comment "Hard disk dedicated to backup has 300GB and other
disk has 245gb"
        length 117760 mbytes  ## i.e. 115gb per tape
}

define dumptype hard-disk-dump {
        comment "Back up to hard disk instead of tape - using dump"
        holdingdisk no
        options compress-fast, index
        priority high
}

## all disks in disklist use this dumptype:
define dumptype hard-disk-tar {
        hard-disk-dump
        comment "Back up to hard disk instead of tape - using tar"
        program "GNUTAR"
}
---------------------------------------------------

I'm trying to figure out what the optimal backup cycle would be. I've
tried doing just once weekly full dump using 2 tapes per dump with no
incremental dumps so that I have at least one full copy for disaster
recovery if the latest backup failed. Space appeared to be an issue
so I also tried using 3 tapes per dump (and if that worked then we
could get another 300GB hard drive, which is all the space we have
left for, so we can have 6 virtual tapes altogether and use 3 tapes
per dump). However, I get this taper error:
"writing file: No space left on device"
even though there is still space on the virtual tape being written to.
Here is the exact output:

----------------
  taper: tape slot-4 kb 120586208 fm 6 writing file: No space left on
device
  taper: retrying pc1:/home.0 on new tape: [writing file: No space
left on device]
  taper: tape slot-1 kb 120586208 fm 3 writing file: No space left on
device
  taper: retrying pc2:/home.0 on new tape: [writing file: No space
left on device]
  taper: tape slot-2 kb 4284096 fm 1 [OK]
---------------

Is error for slot-4 and slot-1 because AMANDA has to fit the entire
dump of one of the entries in the disklist to one tape? But why is
slot-2 not filled?

Maybe my amanda.conf is all wrong? Here is what I have:

--------- snipped from amanda.conf ----------------

inparallel 4            # maximum dumpers that will run in parallel
# half of 100mps (seeing as it will be run during slow hours
netusage  100 mbps

# a filesystem is due for a full backup once every <dumpcycle> days
dumpcycle 1 week        # the number of days in the normal dump cycle
tapecycle 4 tapes       # the number of active tapes in rotation
runspercycle 0          # the number of amdump runs in dumpcycle days
                        # (0 means identical to dumpcycle)

bumpsize 20 MB          # minimum savings (threshold) to bump level 1
-> 2
bumpdays     1          # minimum days at each level
bumpmult     4          # threshold = bumpsize * (level-1)**bumpmult

etimeout 300            # seconds per filesystem for estimates

runtapes     3          # number of tapes to be used in a single run
of amdump
tpchanger "chg-multi"
changerfile "/etc/amanda/weekly/changer.conf"
## tapedev & changerdev is not parsed by chg-multi


autoflush on
tapetype HARD-DISK
---------------------- snipped ---------------------

Sorry for the lengthy question -- any help will be most
appreciated!!!!!


g