> -----Original Message-----
> From: [EMAIL PROTECTED]
> [mailto:[EMAIL PROTECTED]]On Behalf Of
> Ray Olszewski
> Sent: Wednesday, February 19, 2003 11:43 AM
> To: [EMAIL PROTECTED]; [EMAIL PROTECTED]
> Subject: Re: [leaf-user] Cable modem ip change hangs system
>
>
> At 08:12 AM 2/19/03 -0500, Jim Hubbard wrote:
> >I've got 2 Bering 1.0 boxes (different locations) with the 2.4.20
> >kernel, each using a 3nic system.  My isp is Roadrunner.
> 3 times now,
> >Roadrunner service has died briefly and when it came back
> up my Bering
> >box can't connect.  Logs show that pump cannot renew the
> lease.  Tried
> >restarting pump, the network, and the nic, but still nothing.  Each
> >time this has happened, a reboot fixes everything and I
> notice that my
> >ip address has changed.
>
> It's a bit hard to tell from this description what exactly
> is going on ...
> but have you told Shorewall that you are using DHCP to get
> your external
> address? The symptoms you describe are consistent with
> Shorewall blocking
> DHCP ... a failed attempt to get a new lease causes punp to
> *try* to keep
> using its old lease address, and that works except when RR
> decides it needs
> to change.
>

Sorry for being brief, I was hoping others had already seen this
problem.
Here's my /etc/shorewall/interfaces:
net     eth0            detect          dhcp,routefilter,norfc1918
loc     eth1            detect          dhcp,routestopped
dmz     eth2            detect          dhcp,routestopped

I don't think this is a shorewall problem.  There's nothing in the
logs to indicate any problem other than pump failing to renew the
lease.


> If that guess isn't it ... and information you have but did
> not report here
> could easily rule out this possibility ... perhaps a few
> more details would
> help.
>
> When you say "Roadrunner service has died briefly", what do
> you mean? The
> line fails? The server at the other end fails? What kind of service
> interruption are you seeing?

As I said I have two Bering boxes (home and work, both in the same
town).  The first time this happened, Roadrunner had a widespread
outage and the "cable" light went out for several hours on both
systems.  The next time it happened was last night at home, everything
was fine one minute and then I noticed the "cable" light flicker
briefly for a few seconds and then go back on again.


>
> After it comes back and "my Bering box can't connect", what
> does the
> router's configuration look like? In particular, is the
> router still trying
> to use the "old" external IP address, or is the external interface
> unconfigured (no IP address)?

In each case, the external interface is up (it's listed when I issue
"ip addr") but has no ip address, and pump is trying over and over to
renew the lease, but failing.  I wish I had saved the exact log
message, but it seemed to indicate that Roadrunner's DHCP thought my
MAC address was incorrect.


>
> Finally, after a service interruption, why does the Bering
> box *try* to get
> a new DHCP lease in the first place? If the old lease has
> not expired, it
> should believe that it has a valid IP address (and the
> server at the other
> end should too -- in that respect, this sounds like a case
> where RR is
> using DHCP but not following the rules for using it right).
>

I don't know that it's a problem with pump, and I don't know why it's
trying to renew the lease unless the Roadrunner DHCP service told it
the lease had expired.  BTW, I think I'm right in saying that the
Roadrunner DHCP "server" is actually just a function of the cable
modem itself, not some central computer.  Anyway, in each case pump is
the first mention of an error in the logs, and it keeps trying but
failing to renew.


> >Only thing I could think of is to remove the "nodns" parameter from
> >pump.conf so pump can write resolv.conf, but I don't know
> if that's my
> >problem or not.  My isp's nameserver ip's have not changed.
>
> Improbable, based on your description.

I agree.

>
> >During
> >normal operation, logs show pump renews the lease fine (2
> or 3 times a
> >day at different hours for some reason) it just seems to have a
> >problem when the ip changes.  Any ideas?
>
> Does it have a problem "when the ip changes" independent of
> these "service
> has died briefly" settings?

I've never noticed an ip change while running Bering except after
outages as I've described.  Up until the last month or so, which
happens to coincide with the time I started using Bering, my ip
address almost never changed.  Before Bering, I was using
Redhat/Shorewall as my router OS (on the same box at home) and I never
had a problem with ip changes, which used to occur only after serveral
months.  What I can't tell for sure then, is whether Roadrunner is
acting flakey or Bering is.  If Roadrunner was the culprit though, why
would a Bering reboot fix the problem?  I think I will try unplugging
the power from the cable modem for a while to see if I can force an ip
change and see how Bering reacts.

>
>
> Ray Olszewski                                 -- Han Solo


Thanks for your time, and I will appreciate any more ideas.  I was
just hoping someone had seen this behaviour before.

-Jim



-------------------------------------------------------
This SF.net email is sponsored by: SlickEdit Inc. Develop an edge.
The most comprehensive and flexible code editor you can use.
Code faster. C/C++, C#, Java, HTML, XML, many more. FREE 30-Day Trial.
www.slickedit.com/sourceforge
------------------------------------------------------------------------
leaf-user mailing list: [EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/leaf-user
SR FAQ: http://leaf-project.org/pub/doc/docmanager/docid_1891.html

Reply via email to