[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Modem slowdown after hamm upgrade?



I too have noticed a ppp slow down on my bo system.  A while back I
updated a bunch of packages, and I think ppp was one of them.  Here is
what I think I know about that, even though I dont understand it. 

I think that /etc/ppp/options used to have:
mru 542
That line is commented out in the most recent bo ppp package.

I tried to uncommenting 'mru 542' in /etc/ppp/options, but cant establish
a ppp link with that setting. 

I know that mru has been setable in the past and a setting of 542 has
given much better performance for me than the default of 1500.

I dont know if my problem is related to yours, but both look possibly ppp
version related.  

Mike


On Tue, 30 Jun 1998, Simon Blake wrote:

> Hi there
> 
> I've a dual PPro unit that has been a 33K6 PPP PAP dialin server for the
> last year (hamm for the last 3 months), and worked fine, so long as I ran
> pppd v2.2 - I kept the old binary lying around and just symlinked it back
> in after every ppp update from hamm, because the v2.3 binary didn't work,
> and I couldn't be bothered working out why :-).
> 
> Since installing the hamm beta a few days back, my PPP performance has
> plummeted - everything still works, but I get ~1Kbps rather than the
> ~33Kbps I'd normally see.  Since then I've tried 2.0.34 with ppp 2.2, and
> 2.1.107 with ppp 2.3.5, to no avail.  Everything seems to work fine (users
> authenticate OK etc), just deathly slow.  There are two modems in the box,
> showing identical behaviour (both are setserial'd to spd_vhi, and both
> are set to 115200 in the ppp options file), and a 14K4 in another hamm
> machine shows similar poor performance. 
> 
> >From my (hamm) box at home, with a 56K modem, connected to a local 56K
> ISP, I get (running 'pppstats -w 10' while an FTP is going on):
> 
>       IN   PACK VJCOMP  VJUNC  VJERR  |      OUT   PACK VJCOMP  VJUNC
>    47905     35     35      0      0  |      268     19     18      0
>    47981     36     35      0      0  |      213     20     19      0
>    49461     38     36      0      0  |      269     19     18      0
>    47981     36     35      0      0  |      258     18     17      0
>    49038     35     35      0      0  |      201     19     18      0
> 
> Same machine dialed into the 33K server at work, with an FTP happening:
> 
>       IN   PACK VJCOMP  VJUNC  VJERR  |      OUT   PACK VJCOMP  VJUNC
>     1029      3      2      0      0  |       94      2      1      0
>     1017      2      2      0      0  |      120      4      2      0
>        0      0      0      0      0  |        0      0      0      0
>      509      1      1      0      0  |       94      2      1      0
>      508      1      1      0      0  |       26      2      1      0
> 
> I have no idea what is going on here - that's a 10 second gap when nothing
> happened at all!  I've also run a test with minicom and sz, and that
> showed similar performance - around the ~1Kbps mark, so I'm pretty
> confidant the problem doesn't lie with ppp, rather that something has
> changed in the serial port or mgetty setup in the new system?  
> 
> Any ideas welcome - my users are about to send out the lynch mob, and the
> idea of moving the modems back to a bo box doesn't really appeal.
> 
> TIA
> 
> Cheers
> Si


--  
To UNSUBSCRIBE, email to debian-user-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


Reply to: