[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [linux-tr] madge problems
On this same issue, is there anyway that you could look at the souce code
and possibly put in an option to turn off source routing completely (similar
to the one available for IBM Token-Ring cards)? We are in a 100%
non-source-routed environment, and we have the need for PCI speeds, and, as
I understand it, the IBM module does not support PCI cards.
Also, do you think the Madge module could be recompiled under 2.2 and work?
I have successfully compiled the Madge module under 2.0.35 (it was developed
under 2.0.26, I beleive), but with source-routing support only.
If this is not your area, could you please point me to someone who might be
able to help?
Thanks much, in advance,
Great American Insurance Company
From: Paul Norton <email@example.com>
To: firstname.lastname@example.org <email@example.com>
Date: Thursday, April 01, 1999 12:32 PM
Subject: Re: [linux-tr] madge problems
>The default s-r timeout is 10 minutes and is configurable through the
>/proc/sys/net/token-ring/rif_timeout variable. The timeout value for
>each cache entry is reset on every access, whether it's sending a
>frame or reciving one. So the cache entry for on ongoing, active
>connection doesn't expire unless there is a 10 minute timespan between
>frames. That's the theory anyway. :-)
>You can see the cache and the expiration times for each entry through
>the /proc/net/tr_rif file. If it's not working as designed I'd like
>to know about it.
>I have a bm-2 adapter as well as the source to the Madge driver.
>Unfortunately I'm currently trying to debug another problem that
>requires an IBM adapter. I even have my bridge 95% set up, so I'll
>soon have a source-routing network at home to test with.
>Alfred Reed Collins writes:
> > I experienced a problem with similar symptoms, but with a different
> > adapter. The problem turned out to be source-route timeout. When the
> > machine first boots, and makes outbound connections, it works for about
> > minutes, after which the source-route entries expire and the interface
> > stops working until reboot.
> > ****