Stefan,
Thank you, yes, I do have inparallel, set the holdingdisk size and
spindle numbers default (to -1 ?) so they shouldn't be an issue.
I think Paul had it right, I'd missed "maxdumps", I'm sure I'd
have spotted it if present as "1" but I'm running a version of
amanda.conf that doesn't include it.
We tend to carry amanda.conf forward from one version to another
even when upgrading the software itself. Probably a bad policy, I
mean it prevents typos but we end up losing out on the newer parameters
until we hit a call and have to go digging.
thanks,
Brian
[Charset ISO-8859-15 unsupported, filtering to ASCII...]
> Hi, Brian,
>
> on Mittwoch, 21. J_nner 2004 at 16:57 you wrote to amanda-users:
>
> BC> My question is why amanda isn't allowing concurrently with the larger
> BC> partitions when I'm certain we have enough spool space.
>
> BC> How do I do about go about finding out why these partitions aren't
> BC> dumping at the same time, actually, its just an assumption that they
> BC> aren't but given the run time of the partitions and elapse time of
> BC> the job its probably a pretty good guess.
>
> BC> We have dumpers 4, I can include as much of the amanda.conf file as
> BC> anyone wants.
>
> Have you TOLD AMANDA about the doubled space she can use?
> (use-parameter in the holdingdisk-section).
>
> If you use spindle-numbers in your disklist make sure you don`t set
> the same for the both DLEs you want to parallelize.
>
> BC> Run Time (hrs:min) 10:58
> BC> Dump Time (hrs:min) 10:18
>
> There isn_t much parallelism going on, as these numbers tell us.
>
> Let us see your disklist and the corresponding dumptypes as well as
> your parameter "inparallel", and the holding-disk section(s) for a
> start.
>
> --
> best regards,
> Stefan
>
> Stefan G. Weichinger
> mailto:[EMAIL PROTECTED]
>
>
>
