Update - Re: OBSD 2.8 and Bridge problems 2nd post.

[Louis Hopcraft <lhopcraft@telus.net>]

Thanks to those who responded.  

Here's an update to my problem and resolution.

I ended up having to make a trip to the customer site, to deal with this
problem.  Upon arrival I noticed that they were using a 10/100 auto
switch brand name "netbrand" (Which I've never heard of).  Bottom line
is this.  It turns out that the switch has a couple of bad ports which
in turn had alot of col's.  I "disabled" those ports and told them that
they should get a better quality switch.

Thanks
Louis.

On 01 Aug 2001 10:57:17 -0700, Louis Hopcraft wrote:
> Hi all,  Sorry about the long post, but I posted this message a few
> day's ago and I'm still have a few problems tracking down the problem.
> The work stations are plugged into a hub just for them which in turn
> goes into ep0.  The servers are also plugged in a hub just for them and
> they are plugged into rl1, and rl0 is plugged into a 100MB connection.
> The problem I'm having here is two pc's connected to rl1 (they are also
> the most busy).  Users with work stations need to connect to these two
> from behind the firewall (ep0)
> half the time they are pingable from behind ep0 and the other half they
> are not.  When they are not pingable, and I run a tcpdump on rl1 (where
> the servers are connected, with the bridge) I can see the icmp requests
> and reply, but from the workstations they are still timing out.  It is
> only happing to these 2 servers, and all of the other 15 or so are fine.
> This is all I have in my ipf.rules file.
> 
> pass  in quick on rl1 proto tcp from any to any 
> pass  in quick on rl1 proto udp from any to any 
> pass  in quick on rl1 proto icmp from any to any
> #block in quick on rl1 # nuh-uh, we're only passing tcp/udp/icmp sir. 
> 
> pass  in quick on ep0 proto tcp from any to any
> pass  in quick on ep0 proto udp from any to any
> pass  in quick on ep0 proto icmp from any to any
> pass out quick on ep0 proto icmp from any to any
> 
> As I had said in my orginal post, At first I can ping the 2 servers in
> question, but after 8 or 15 hours they are unreachable from behind ep0
> (internal network) and with-in an hour or so they'll can back and then
> the whole process repeats, and during this time all of the other servers
> are still reachable from behind ep0 when these 2 troubled ones are not,
> and they are all connected thur the same hub.
> 
> Thanks again
> Louis.
> 
> 
> 
> 
> >I'm having a weird problem.  I'm running an OBSD 2.8 box running as a
> >Bridge with three network cards.  rl0 & rl1 are bridged and ep0 is a
> >internal ipaddy range behind the firewall.  I have 15 routeable ip's on
> >rl1 and I can get to them from outside and internaly fine, with the
> >exception of one box.  The one in question here I can ping from the
> >outside and I can also connect to it as well.  But from the internal
> >network I cannot ping or connect.  Does anyone have any ideas of where
> >I
> >should look for this problem?  I've reboot the box in question and
> >checked it and it looks all ok.  I've checked my ipf.rules and (they
> >are
> >pretty lean at this stage) they look fine as well.