Subject: Re: [netatalk-admins] Re: Problems w/ FreeBSD 3.1 as seeding router?
From: Douglas E. Wegscheid (wegscd@whirlpool.com)
Date: Tue May 11 1999 - 08:34:41 EDT
At 02:06 PM 5/7/1999 -0700, a sun wrote:
>in any case, i thought that this was a situation where nothing showed
>up in the chooser even in the routerless case. is that not correct?
>usually, that's a multicast problem. in addition, there used to be a
>problem with net.0 queries in linux, but that usually manifests in
>situations where you're going through routers.
well, I did more work this weekend, and that's where it is. Macs are still
not seeing the FreeBSD 3.1 with the freebsd.org compiled asun 2.1.3, either
without a atalkd.conf (coming up zoneless), or with a
le0 -router -phase 1 -net 100-200 -addr 100.2 -zone Foo
I can take a netbsd-1.3.2-mac68k system, configure it identically (either
no atalkd.conf, or with -router), and all is well (the Macs see it). I am
not running both boxes at the same time, and I am shutting all Macs down on
the wire in between switching boxes.
My Appletalk knowledge is about nil, but I did some work with the sniffer;
when the mac comes up, it does a series of probes (to make sure it has a
unique address?), then sends a ZIP packet for 'info request'. NetBSD
running -router will answer, FreeBSD running -router will not. The
destination MAC address in the packet is the Apple broadcase address (I
didn't switch to numeric mode to see what that was).
Is there a known bug in FreeBSD-3.1 release receiving Appletalk multicasts
(broadcasts)? Is there some debugging switches or DEFINES I can throw onto
atalkd so I can see if it is getting ZIP packets or not? That would help
narrow it down to atalkd is getting the packets and ignoring them, atalkd
is never getting the packets, or atalkd is getting the packets, sending
them out, and there aren't getting to the wire? What more information can I
collect?
I haven't tested out bringing up the NetBSD system as -router, letting
FreeBSD come up and discover the router, and see what happens then. That's
tonight's exercise...
-
Douglas E. Wegscheid
working for, but not expressing the position of, Whirlpool Corporation
wegscd@whirlpool.com
-
A wrong note played hesitatingly is a wrong note. A wrong note played with
conviction is interpretation.
This archive was generated by hypermail 2b28 : Sat Dec 18 1999 - 16:16:42 EST