Ascend Archive
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Fwd: Re: (ASCEND) Intermittent loss routes?]



-- BEGIN included message

Alex P wrote:
> 
> There's also a big chance that it's the 'ol USR Sportster bugs showing up
> again.
> 
> Alex P
> 
> On Tue, 19 May 1998, Ricardo Freire wrote:
> 
> > Chen,
> >
> > I think there's a BIG chance that your customers have noise in THEIR
> > analogic lines.
> > In such cases, the connection behavior looks exactely like you described.
> > But, what causes those "lags"?
> >
> > 1) modems' fallback/fall forward: when connection conditions change due to a
> > noise, some modems can re-negociate the connection speed. It longs for 15
> > seconds, or more.
> >
> > 2) modems retraining: when more severe conditions occur, the modems can
> > reinitiate ALL the connection, from handshaking! It takes 40-60 seconds, and
> > in the meanwhile NO data traffic is possible!
> >
> > 3) The third case is simply the disconnection, caused by some noise level
> > that the modems cannot handle.
> >
> > Your customer can hear events (1) and (2): in modem
> > settings/connections/advanced/extra settings, they can put M3, which is the
> > modem command to make buzzer stay up. When, for example, retraining occurs,
> > they will be able to hear the SAME sounds they hear everyday, when
> > connecting normally.
> >
> > So, they have to contact local teclo, to fix their lines.
> >
> > Regards,
> >
> > Ricardo
> >
> >
> > -----Original Message-----
> > From: Stephen Chen <stephen@eni.net>
> > To: ascend-users@max.bungi.com <ascend-users@max.bungi.com>
> > Date: Terça-feira, 19 de Maio de 1998 15:05
> > Subject: (ASCEND) Intermittent loss routes?
> >
> >
> > >Some of our dialup customers have complained about losing route to the net
> > >after connected.  The problem is intermittent, the user can login and
> > >access the net for certain amount of time, then unable to access the net,
> > >and later on regain the connectivity, all within the same call.  Traceroute
> > >(while having problems) shows extremely slow time (or timeouts) between the
> > >MAX and user's dynamically assigned IP.  And 'modemdiag' shows the line
> > >quality to be 20-40 (excellent).
> > >
> > >Is this some type of routing problem, or something more serious?
> > >
> > >The MAX are 4004 running either 6.0.0 or 6.0.2, with average of 3 PRI lines
> > >and 6 K56-mod12 cards.  MAX is not running OSPF, with a static route to the
> > >gateway.
> > >
> > >any help is greatly appreciated.
> > >
> > >Regards,
> > >
> > >Stephen
> > >Epoch Internet
> > >++ Ascend Users Mailing List ++
> > >To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
> > >To get FAQ'd: <http://www.nealis.net/ascend/faq>
> >
> > ++ Ascend Users Mailing List ++
> > To unsubscribe:       send unsubscribe to ascend-users-request@bungi.com
> > To get FAQ'd: <http://www.nealis.net/ascend/faq>
> >
> 
> ++ Ascend Users Mailing List ++
> To unsubscribe: send unsubscribe to ascend-users-request@bungi.com
> To get FAQ'd:   <http://www.nealis.net/ascend/faq>

We have seen this same problem with USR Sportsters.  This is the x2
version of Spiral Death Syndrome. The fix in most cases is to add
S15=128  to the init string to disable V.42 hardware error correction.
It may also be necessary to  add &K0 to the init string to disable
V.42bis hardware data compression, though this should be a last resort,
as performance will suffer when downloading non-compressed files. 

Bill
wheels@gloryroad.net

++ Ascend Users Mailing List ++
To unsubscribe:	send unsubscribe to ascend-users-request@bungi.com
To get FAQ'd:	<http://www.nealis.net/ascend/faq>

-- END included message