The sendsize.debug on the failing machine does not show a device being
userd. I examined the same file on another machine that is working, and it
shows "DUMP: Dumping /dev/sda1 (/(dir var)) to standard output. The
problem maching is running FreeBSD 4.2 and the agreeable maching is
running Debian 2.2. Should I update the Dump? The disklist entries are
identical, except for hostname.
On Thu, 5 Apr 2001, John R. Jackson wrote:
> >sendsize.xxxxxxxxx.debug shows that all dumps are failing with:
> >
> >DUMP: bad sblock magic number
> >DUMP: The ENTIRE dump is reported
> >
> >30 Gb partition. Is there a 2 Gb partition size limit? ...
>
> There may be, but I doubt it is the problem.
>
> >Any idea what might be causing this?
>
> It is not pointing dump at the right place. For instance, dump is trying
> to process a file/directory instead of a device.
>
> Or the dump it is trying to run does not understand the file system type,
> for instance, running ufsdump on vxfs.
>
> Take a look at the dump command Amanda actually ran (also listed in the
> sendsize*debug file) and make sure it is right.
>
> John R. Jackson, Technical Software Specialist, [EMAIL PROTECTED]
>
