Re: Socket problems with latest -current

If memory serves me right, Doug Barton wrote:
> Bruce A. Mah wrote:

>> I'm trying to reproduce this problem but without success.  I did an
>> install of 6.2-RC1/i386 from CD into a Parallels VM, then manually
>> updated /etc/rc.d/auto_linklocal to version 1.1.2.3.2.1 (same as 1.4 you
>> tested with). 
> 
> I'm not sure if they're relevant or not, but the differences between
> what you did and what I did are:
> 
> 1. I installed on a real box
> 2. I installed 6.1-RELEASE from CD, then cvsup'ed to RELENG_6.
> 
> One other possible difference, I'm on a Core 2 Duo, running i386 SMP.
> 
> You might also try enabling named for your testing. The default
> configuration should give you a simple local resolver which you can
> then test things like 'dig _at_127.0.0.1 www.yahoo.com'

Thanks for the info.  BIND seems to work OK in this environment too.
(What was the failure mode for you?)

Part of me is hoping that there's something odd in your environment,
especially because I haven't heard any other similar reports of problems
so far on current_at_, stable_at_, or net_at_.  I figure if sendmail or named
can't bind sockets there'd be a pretty loud outcry from a lot of people.

BTW, I started looking at this because re_at_ needs to decide what to do
here for 6.2.  If a default (or even "normal") configuration is possible
to produce the problems you saw, we obviously can't ship this way.

Would it be too much to ask for a (possibly slightly sanitized)
/etc/rc.conf and kernel configuration file from the machine in question?
 (Privately obviously.)

Oh yeah.  Is it possible that the machine has a hostname that only has a
AAAA record (no A record) in DNS or some other thing like this?  I don't
expect that you'd do something so "interesting" but I had to ask.

Thanks!

Bruce.

PS.  I'm in the process of updating a physical machine to the tip of
RELENG_6 but this takes awhile (it's a UP 1GHz P-III and believe it or
not the fastest FreeBSD machine I own).