On Wed, Sep 22, 2004 at 05:30:11PM -0500, Jason Miller wrote:
> I edited [the tapetype entry]
> to see if it made a difference and of course it didn't.

>From amanda(8):
> Note that [tapetype's "length"] value is only used by Amanda to schedule
> which  backups  will  be  run.  Once the backups start,
> Amanda will continue to write to a tape until  it  gets
> an  error,  regardless  of  what  value  is entered for
> length (but see the OUTPUT DRIVERS  section  later  for
> exceptions).

The same is true of "filemark" and, I believe, "speed".

As Jon pointed out:
> It wrote 1 file (fm 1) and about 448KB "successfully".
> [...] Then it hit an error.

There's no point messing with the tapetype entry until you've
solved the underlying problem of only being able to write 0.5 MB
to the tape.

> That file was the amanda tape header, not a backup dump.

As an aside, it must have been the header (32 KB + 1 file mark),
plus the first part of the first dump (416 KB, presumably).

If you haven't already:
  - try a different tape -- maybe the one you've been testing
    with is damaged

  - look at the Amanda debugging files (in /tmp/amanda by
    default, but you can put them elsewhere with a "configure"
    option)

  - look at any relevent kernel messages

  - try writing to the tape by hand (using "dd bs=32k" to match
    Amanda's usage of the drive).  By eliminating Amanda from the
    equation:
      - you can determine whether the problem is with Amanda or,
        as I strongly suspect, at a lower level
      - if the latter, you get any error messages directly,
        neither "interpreted" by Amanda nor shunted off to the
        /tmp/amanda logs

--

|  | /\
|-_|/  >   Eric Siegerman, Toronto, Ont.        erics AT telepres DOT com
|  |  /
The animal that coils in a circle is the serpent; that's why so
many cults and myths of the serpent exist, because it's hard to
represent the return of the sun by the coiling of a hippopotamus.
        - Umberto Eco, "Foucault's Pendulum"