From: Mark Lasarko (mlasarko@co.ba.md.us)
Date: Wed Feb 22 2006 - 09:07:29 GMT-3
Not knowing the complete scope I can share that I have seen *small*
networks with thousands of routes and *large* networks with only dozens.
If you plan out your address space well the statics are probably not such
a big deal, you summarize; have your primary plus a floater for sake of
redundancy. If the address space is not contiguous - badly fragmented;
Then I would say you are likely better off maintaining an active IGP
(OSPF, in your case) to deal with the situation.
In short aggregation and summarization would be the keys to me to
"address" the internal issue(s) - no pun intended.
Externally, what other requirements are calling for BGP everywhere?
Are there other services that mandate BGP beyond the IPS edge routers?
I would look very closely to see if you *require* OSPF and BGP on *every* core
router.
HTH,
~M
>>> "Jens Petter Eikeland" <jenseike@start.no> 02/22/06 6:10 AM >>>
Can anybody try give me some good arguments for running bgp over ospf
internaly compared to running bgp over static routes.. We are going soon to
design redundancy in to our net and I am trying to tell that this would work
best with running dynamic routing in under BGP. There is a guy that are
trying to convince me that running only BGP would be better (with static
routes). We have this today. I am trying to say that this is not a good
solution. We have over 200 static routes on our core router, and it is not
very easy to keep track of anything... Anyway, when we are going over to
both have internal redundancy (internally)and external redundancy to our ISP
how could this possible be scalable with static routing. We are an ISP, and
have got many tousen customers that we are going to get in to to net soon.
This will generate more static routes... My suggestion is that we run OSPF
internally with BGP on every router in out core net, but he is saying that
keeping static routes and BGP is still the best solution..
Looking for some good arguments
Jens
This archive was generated by hypermail 2.1.4 : Wed Mar 01 2006 - 11:28:18 GMT-3