|
Hi Joe!
On Thu, 17 Jul 2003, Joseph Mack wrote:
> Stephan Wonczak wrote:
> >
> > > > taken over service from machine A, but both are still running. This
> > > > happens e.g. during maintenance. Machine B has both :0 and :1-Adresses,
> > > > machine A does no longer (verifiable by ifconfig).
>
> what does `ip` show?
Nothing much different or even interesting:
Machine A (lvr7, currently director)
[root@lvr7 root]# ip addr show eth2 4: eth2: <BROADCAST,MULTICAST,UP> mtu
1500 qdisc pfifo_fast qlen 100
link/ether 00:02:b3:be:8a:9e brd ff:ff:ff:ff:ff:ff
inet 134.95.19.105/24 brd 134.95.19.255 scope global eth2
inet 192.168.19.7/24 brd 192.168.19.255 scope global eth2:3
inet 134.95.19.15/24 brd 134.95.19.255 scope global secondary eth2:0
inet 192.168.19.254/24 brd 192.168.19.255 scope global secondary eth2:3:0
inet 134.95.19.54/24 brd 134.95.19.255 scope global secondary eth2:1
Machine B (lvr8, had the director for a short time, *and* was working
well during the short time it had the director function)
[root@lvr8 root]# ip addr show eth2 4: eth2: <BROADCAST,MULTICAST,UP> mtu
1500 qdisc pfifo_fast qlen 100
link/ether 00:02:b3:be:8a:a6 brd ff:ff:ff:ff:ff:ff
inet 134.95.19.106/24 brd 134.95.19.255 scope global eth2
inet 192.168.19.8/24 brd 192.168.19.255 scope global eth2:3
inet 134.95.19.111/24 brd 134.95.19.255 scope global secondary eth2:4
inet 134.95.19.58/24 brd 134.95.19.255 scope global secondary eth2:0
For explanation:
134.95.19.105 Real IP of machine A (lvr7)
134.95.19.106 Real IP of machine B (lvr8)
192.168.19.254 NAT-Router address
134.95.19.15 Virtual IP of service A (lrv)
134.95.19.54 Virtual IP of service B (msa)
The real nodes are in the 192.168.19.xxx-net.
> If addresses are being installed and removed by ip, then
> ifconfig won't see them
Oh, I wasn't aware of that. Anyway, this does not seem to be the
problem.
I have just reproduced the problem again (we got around it last time by
rebooting lvr8). Here is a snippet of 'tcpdump -e -i eth2 -l arp' from a
time after the director is back on lvr7:
16:10:15.916326 0:0:be:a6:50:21 Broadcast arp 60: arp who-has
lrv.rrz.uni-koeln.de (Broadcast) tell starfire.rrz.Uni-Koeln.DE
16:10:15.916360 0:2:b3:be:8a:a6 0:0:be:a6:50:21 arp 42: arp reply
lrv.rrz.uni-koeln.de is-at 0:2:b3:be:8a:a6
Whoops! 0:2:b3:be:8a:a6 is lvr8! Now the big question: Why does lvr8
insist on answering? The mind boggles :-) (nb: I may have mentioned that I
am none too firm with the internals of networking. So any suggestions
are welcome, even if they may seem too trivial to you)
Thanks for your help!
Dipl. Chem. Dr. Stephan Wonczak
Institut fuer Angewandte Informatik (ZAIK)
Regionales Rechenzentrum der Universitaet zu Koeln (RRZK)
Universitaet zu Koeln, Robert-Koch-Strasse 10, 50931 Koeln
Tel: ++49/(0)221/478-5577, Fax: ++49/(0)221/478-5590
|
