I can get 30MB/sec transfer using /dev/zero but it would make sense to use
some sort of random data (no /dev/random in AIX) for a more realistic
result.

I didn't use the timer from dd as it includes it takes for tape negotiation
and rewind (I noted it from a throughput indicator from our Fibre-Switch
display)

Suad
--
On Fri, Apr 20, 2001 at 03:42:05PM -0500, S W Branch wrote:
On Fri, Apr 20, 2001 at 03:42:05PM -0500, S W Branch wrote:
> >timex dd if=/dev/zero of=/dev/rmt1 bs=1024 count=1000
> >
> >The result was real 21.06 seconds.  This falls way short of what I would
> >expect when the advertised sustained transfer rate is supposed to be
> >15MB/sec native (30MB/sec compressed). ...
>
> 1K seems like a very small block size. Did you try other block sizes to see
> if that made any difference in speed?  I think I remember some discussion
> quite some time ago about what blocksize TSM used - the AIX setting for the
> tape device or some other value - I think the answer was that it used a 64K
> blocksize.