ospf and static routes

Unanswered Question
Jan 28th, 2010

I'm trying to convert our network to dynamic routes using ospf, as it's all static routes at the moment (in excess of 600 devices, all statically routed!!!!).

Anyway, I'm having to keep service up whilst I'm doing this so I am going to have to have both ospf and static routes running at the same time. My question is this:

If I redistribute the static routes in to ospf, will the network still use the static routes (as the administrative distance is lower for a static route)? If so, would it be better to just run ospf and get rid of the statics in order to avoid problems (I suspect the answer is yes, and I don't mind doing this...I'll just need to get ospf to learn its routes then remove the statics one at a time to see if there are any problems....which there shouldn't be if ospf has learned the route).

Thanks

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Giuseppe Larosa Thu, 01/28/2010 - 02:17

Hello Codflangers,

if you use redistributes static subnets within process ospf, static routes installed in routing table are introduced in the OSPF database and propagated to other OSPF speakers.

In this sense they can coexist but a full mesh of static routes means that there can be redundant that is the same static route for prefix A can be defined on multiple devices and all of them would try to inject external OSPF routes for it.

A possible approach can be that of designing a multi area OSPF domain with a backbone area and to migrate to OSPF devices of a specific non zero area.

A check of IP OSPF database on ABR (routers between area 0 and area X) can allow to verify the completeness.

The two ABR can generate inter area summary routes that represent all of prefixes included in network area commands

Because sh ip route A will be still the static route you need to use

sh ip ospf database external A

to verify presence of the prefix.

There is also the possibilty that some prefixes will be seen as internal if

Once all prefixes are present in DB you can start to remove static routes in the area devices until all static routes are removed, or at least all those that can be removed.

Hope to help

Giuseppe

Jon Marshall Thu, 01/28/2010 - 04:27

codflanglers wrote:

I'm trying to convert our network to dynamic routes using ospf, as it's all static routes at the moment (in excess of 600 devices, all statically routed!!!!).

Anyway, I'm having to keep service up whilst I'm doing this so I am going to have to have both ospf and static routes running at the same time. My question is this:

If I redistribute the static routes in to ospf, will the network still use the static routes (as the administrative distance is lower for a static route)? If so, would it be better to just run ospf and get rid of the statics in order to avoid problems (I suspect the answer is yes, and I don't mind doing this...I'll just need to get ospf to learn its routes then remove the statics one at a time to see if there are any problems....which there shouldn't be if ospf has learned the route).

Thanks

I agree with Guseppe and you on this. It would be much better to simply add OSPF and check that OSPF is learning all the correct routes and then simply remove the statics. If you redistribute the statics into OSPF that is a recipe for disaster as you OSFP will then have all the routes but it could be easy to make a mistake, remove a static and then realise that OSPF has lost the route as well.

If you configure OSPF with the correct network statements etc. and no redistribution then you can be sure if you are seeing the route you can safely remove the static.

Jon

Actions

This Discussion