Date: Wed, 24 Dec 1997 19:44:16 +0000 From: Brian Somers <brian@awfulhak.org> To: Kwoody <kwoody@citytel.net> Cc: Brian Somers <brian@awfulhak.org>, freebsd-questions <freebsd-questions@freebsd.org> Subject: Re: Kernel messages. Message-ID: <199712241944.TAA24962@awfulhak.demon.co.uk> In-Reply-To: Your message of "Wed, 24 Dec 1997 09:52:11 PST." <Pine.BSF.3.91.971224090933.4118B-100000@mybsd.net>
next in thread | previous in thread | raw e-mail | index | archive | help
>
>
> On Tue, 23 Dec 1997, Brian Somers wrote:
>
> >
> > Nope, the ``File exists'' is the text associated with errno EEXIST.
> > In this context it means that there's already an interface with the
> > given (via `set ifaddr') destination address. The `wrong ifa'
> > messages indicate that there's already an interface with the given
> > (via `set ifaddr') source address.
>
> I would guess this is becuase when ppp dials out during a cron job for
> mail, when there is no more mail I use kill -INT in the script to drop the
> connection, but all the routes remain in place. Is it ok for routes to remain
> afterwards? Or should they be deleted?
Depends. An INT will tell ppp to terminate the current connection if
any (you should really use `pppctl ... close' - it's more `polite' to
the peer). When in -auto, -ddial or -dedicated mode, this will not
result in ppp exiting - therefore your routes will remain (unless
ppp.linkdown says otherwise).
> > Are you trying to run ppp from /etc/ttys without the -direct switch ?
> > This won't work as ppp daemonizes itself and init will then try to
> > respawn ppp. The result is loads of ppp processes - all failing to
> > run, and an unstable system that probably runs out of file descriptors
> > among other things.
>
> No, I run ppp from the command line manually. I think that getty's message
> may have been caused by me banging away on the keyboard in frustration just
> before all the vtty's logged out.
>
> > This sounds as if ppp was using up all the file descriptors. Was
> > there only a single ppp running or was there loads of them (all
> > spawned by init in /etc/ttys) ? Can you reproduce the problem ?
>
> Nope just one. Ive been back at 0820 of ppp for the last week so have not
> retrured to 1215. I know the odd occasion when I notice cron didnt work
> because ppp died and in the ppp.log is a mesasage about bad file
> descriptors. doesnt happen often though.
>
> >
> > I updated ppp on Dec 19 so that it exits in -auto mode if the
> > specified interface addresses are already configured. You may want
> > to get the latest version.
>
> Destination Gateway Flags Refs Use Netif Expire
> default 204.244.99.76 UGSc 0 4 tun0
> 127.0.0.1 127.0.0.1 UH 0 800 lo0
> 192.168 link#1 UC 0 0
> 192.168.0.2 0:c0:f0:b:8f:9b UHLW 1 13315 lo0
> 204.244.99.76 204.244.99.124 UH 1 0 tun0
>
> this is what a netstat always shows after a ppp session done. All routes
> are still in place. Only thing that will change is the dynamically
> assiged IP and sometimes the gateway, depnding onwhich router I hit.
>
> Will this affect the lastest version of PPP when running in -auto mode?
> Or just the first time ppp is run in -auto? Should routes be deleted
> after a ppp session is done?
Nope - only if ppp exits.
>From the sounds of it, you should be running `ppp -background ...'.
In background mode, ppp establishes a connection and then slips into
the background, the parent returning 0 for success. This is the best
thing to use for overnight scripts. You can tell if you got the
connection (and can continue), and when ppp is `closed', it exits.
I haven't seen anything about bad file descriptors myself for about 2
months (since a libalias problem was fixed).
> thanks a bunch Brian.
> Keith.
>
--
Brian <brian@Awfulhak.org>, <brian@FreeBSD.org>, <brian@OpenBSD.org>
<http://www.Awfulhak.org>;
Don't _EVER_ lose your sense of humour....
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199712241944.TAA24962>