Date: Mon, 4 Dec 2000 16:36:46 -0500 (EST) From: Mike Remski <remski@ellacoya.com> To: Brian Somers <brian@Awfulhak.org> Cc: mremski@ellacoya.com, freebsd-stable@FreeBSD.ORG Subject: Re: PPP & named oddity Message-ID: <Pine.GSO.4.02.10012041636370.15753-100000@chukchi.ellacoya.com> In-Reply-To: <200012042046.eB4KkvD90834@hak.lan.Awfulhak.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Thanks, will do. Michael Remski Software Engineer mremski@ellacoya.com 7 Henry Clay Drive Merrimack, NH 03054 +1 603.879.7241 direct +1 603.577.5533 fax www.ellacoya.com On Mon, 4 Dec 2000, Brian Somers wrote: > Sorry for the delay - I've had no connectivity over the weekend... > > I think you need to add some bits to your ppp config to find out > what's going on. I'd suggest: > > ppp.linkup: > > MYADDR: > set log -dns -tcp/ip > > ppp.linkdown: > > MYADDR: > set log +dns +tcp/ip > > The idea is to enable tcp/ip and DNS logging while the link is down > and disable it when the link comes up (it's mostly useless then). > > When ppp gets into the state where it's not bringing the link up, you > should be able to look in ppp.linkup and either see that packets are > being blocked or that packets are not being seen. > > Alternatively, you could be seeing ppp failing to connect to your > ISP, blocking up with packets and waiting for the choked timer to > kick in (default of 2 minutes) and drop the queue in the bit-bucket. > If this is the case, you can do something like ``set choked 20'' to > bring that timeout down to a value that'll make ppp more persistent > in trying to connect. > > > Thanks Brian. I took a look at the faq, and had everything set up the > > way it suggested. netstat -rn showed a default route and a route > > between the two 10. addresses for tun0. My /etc/resolv.conf was > > pointing to the nameserver on the local machine: nameserver 0.0.0.0 > > directive. > > I wasn't clear that the fetchmail is running on the same box that is > > running ppp -auto. I think this indicates outgoing traffic is not > > causing ppp to dial. When it gets into this state it seems like named > > had reaped some stuff out of its cache (time to live stuff). Could > > this (named clearing cache) or an iface clear in ppp.linkdown be causing > > an adverse reaction with the routing table? > > > > It will typically work for 3 or 4 days befre getting into this state. > > > > Again, thanks for the suggestions. > > > > Michael Remski > > Software Engineer > > mremski@ellacoya.com > > 7 Henry Clay Drive > > Merrimack, NH 03054 > > +1 603.879.7241 direct > > +1 603.577.5533 fax > > www.ellacoya.com > > > > On Fri, 1 Dec 2000, Brian Somers wrote: > > > > > Check out http://www.FreeBSD.org/FAQ/ppp.html. Your best bet is to > > > set up a diagnostic socket and attach to it when ppp appears hung. > > > You should be able to ``set log local physical'' and see any incoming > > > and outgoing traffic. If there's no incoming traffic, I'd tend to > > > blame the peer. > > > > > > > Hi all. > > > > I have a small network (3 machines) running on 192.168. addresses, > > > > outside connection is 56K modem. I'm starting both named and userland > > > > ppp at boottime. ppp.conf is basically the one that gets installed, > > > > with the enable dns line commented out and the username, phonenumber and > > > > password set. I then have a cron job that runs fetchmail every 4 hours > > > > or so to pull down mail. After a time (maybe a day or so), ppp never > > > > dials out. If I kill -9 and restart it stays ok for awhile again. > > > > ppp is started with -nat -auto -quiet papchap > > > > named just resolves my three machines. > > > > netstat -rn shows the default route pointing out thru tun0. > > > > ppp.linkdown has an iface clear in it. > > > > > > > > ps ax when this is happening does not show anything unusual about ppp. > > > > kernel has the IPDIVERT/IPFIREWALL and NETGRAPH options in it (I am not > > > > running natd, I'm letting ppp handle that) > > > > > > > > Any suggestions of what to look at greatly appreciated. > > > > > > > > Michael Remski > > > > Software Engineer > > > > mremski@ellacoya.com > > > > 7 Henry Clay Drive > > > > Merrimack, NH 03054 > > > > +1 603.879.7241 direct > > > > +1 603.577.5533 fax > > > > www.ellacoya.com > > -- > Brian <brian@Awfulhak.org> <brian@[uk.]FreeBSD.org> > <http://www.Awfulhak.org>; <brian@[uk.]OpenBSD.org> > Don't _EVER_ lose your sense of humour ! > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.GSO.4.02.10012041636370.15753-100000>