Serial port 9600 versus 19200

[Bill Mayhew]

Xref: neoucom unix-pc.general:4214 comp.sys.att:9320

>I have run some file transfers at both 9600 and 19200 and to be 
>quite honest,  I doubt that you will notice any difference.  I 
>couldn't.
>
>bill gunshannon


It does make a difference on my installation.  I use Unix 3.51 with
the HDB uucp kit.  I have a Trailblazer 2000 attached to my serial
port.  A couple of days ago, several of the news batches I received
were 250K (the IBM binaries group).  The xferstats showed a
throughput of 1400 and 1398 char/sec.

If one examines HDB xferstats for only small files, there is not
much difference between 9600 and 19200 bps.  The xferstats include
a fixed overhead time in dispatching the file.  Of course, this is
fair; if all one has is small files, then the overhead is longer
than the transmission time, so the port speed doesn't matter.

The other mitigating factor is having a fast host on the other end.
Before, I had a vax 750 on the other end of my link.  I could
receive 1000+ char/sec from the vax, but the vax could only take
~650 char/sec from my 3b1.  Now that there is an HP 9000/835 on the
other end, I get ~1400 char/sec service both directions.  I believe
that now my 3b1 is the slower of the pair ( :-) ).  Actually, the
Trailblazer is going to be the limiting factor because on a perfect
line the Trailbalzer has 18000 bps raw rate, but that is reduced to
14400 bps because of the overhead internal to the Trailblazer in
managing the channel.


Bill
-- 

Bill Mayhew  Northeastern Ohio Universities College of Medicine
Rootstown, OH  44272-9995  USA    216-325-2511
wtm at uhura.neoucom.edu   ....!uunet!aablue!neoucom!wtm