On Wed, 5 Jan 2000, Greg Mirsky wrote:
> > > > What I would suggest is to take the OSPF statement out and re-enter
> > it...
> > > > I have had some strange problems with OSPF and this usually seems to
> > take
> > > > care of it.
> > > >
> > >
> > > TA DA! This did it!! Funny thing is, this same problem cropped up on two
> > > routers on the other part of the ring. Oh well, chalk one up to
> > > experience. Cisco bug perhaps? Thanks to everyone who replied! :)
> > Not bug and not Cisco's. This is a _feature_ of OSPF.
> > Standard defines no way to clear LSA database, so, if you change topology,
> > old data could mess OSPF operation, until they are expired (some very
> > long time), or router is rebooted, or OSPF is disabled and re-enabled
> > again.
> [GIM] Let me disagree. If you change topology that you lose your
> neighbor(s) you should purge your SPF from all LSA that originated or
> recieved from that neighbor by means of premature aging them out (setting
> age to 3600). It's in the spec.
Topology changes that lead to _loosing_ neighbours are handled OK.
It is a rearranging of existing topology, when all participating routers
are the same, drives OSPF mad. When you drop and reinsert "router ospf <>"
configuration command (or reboot the router), then you _force_ others
to loose this neighbour.
>
> Regards,
> Greg Mirsky
>
> > I wish that Cisco implement a command to disable OSPF process without
> > removing all OSPF configuration (something like "dlsw-disable" command).
> >
> > >
> > >
> >
> > ---------------------------------------------------------
> > Basil (Vasily) Dolmatov CCIE #5347, CCNP-Security, CCDA
> > East Connection ISP, Moscow, Russia. (http://www.east.ru)
>
---------------------------------------------------------
Basil (Vasily) Dolmatov CCIE #5347, CCNP-Security, CCDA
East Connection ISP, Moscow, Russia. (http://www.east.ru)
This archive was generated by hypermail 2b29 : Sun Aug 04 2002 - 04:12:08 EDT