Before we start we need to come up with a migration plan. A migration plan is very simple, we'll look at it in the next few slides. First off, we verify the OSPF configuration and operation. We need to check that what we've got is actually working. We deploy IS-IS over the entire backbone. We change the administrative distance of OSPF to be higher than that of IS-IS, and that's the key. We check for remnants of OSPF, and then we remove OSPF from the entire backbone, and finally confirm IS-IS operation. So let's look at each in turn. First off, verifying the OSPF configuration. So this is implementing all the good practices that we have discussed elsewhere in this series. For example, next-hop-self for iBGP, so that we do not carry external point-to-point links in OSPF. If we do need to carry external point-to-point links, for example, for monitoring, we carry them in iBGP and tag them with a specific community which is visible to the monitoring system only. We remove surplus OSPF 'network' statements. Older versions of Cisco IOS could only handle 'network' statement configuration. In other words, OSPF ran on the subnet, not on the link. For more modern versions of Cisco IOS, those from 12.4 onwards, and Cisco IOS XR, OSPF is operated on the physical interface instead. So we want to modernize our configuration so that it all performs like this. The slide shows a configuration example if you're still running an iOS prior to 12.4. I wouldn't recommend running such an old piece of software, but realistically some networks still have to do this. If you are, then this is the type of configuration required, as an example. More modern versions of the iOS have the example shown in the slide, where OSPF has been configured on each interface. And this slide shows how the same thing may be configured using Cisco IOS XR Where again, OSPF is activated on each interface that requires the adjacency. If IPv6 has already been deployed, then we need to tidy up the OSPFv3 configuration in the same way as we tidied up OSPFv2. In Cisco IOS the router OSPF configuration should look identical to the IPv6 router OSPF configuration. For IOS XR the router OSPF configuration should look identical to the router OSPFv3 configuration. It's essential for the network operator to compare the two configurations for OSPFv2 and OSPFv3 and make sure they are identical, and also check that the IPv4 adjacencies match the IPv6 adjacencies. We don't want to miss anything out. Once we have done the configuration check on the network we should verify OSPF's operation. Verifying operation is important after any cleanup. We need to make sure that the iBGP peers are stable, that the next hop values are all valid, and that the OSPF routing table is complete. If we've already deployed IPv6, we need to check that OSPFv3 configuration is working the same as the OSPFv2 configuration. So as well as all the adjacencies, we want to compare routing table entries and so forth. Once we've done this we're now in a position to deploy IS-IS over the entire backbone. ISPs we'll deploy IPv6 dual-stack across their infrastructure. Every device running an IPv4 IGP will also require to run an IPv6 IGP. They may choose single-topology IS-IS where the v4 and v6 topology is identical across the whole backbone. This needs care though, as adjacent routers need to have both IPv4 and IPv6 on the link. These days more operators are choosing to use multi-topology IS-IS, where the v4 and v6 topology can differ, and this gives a lot more flexibility for operators who are doing incremental rollouts of IPv6. The IS-IS deployment can happen at any time because, at least on Cisco IOS and most other vendors, the IS-IS protocol distance is higher than that of OSPF's. A higher protocol distance means the protocol has lower priority. So even if IS-IS is up and running, and carrying adjacencies and reachability information, OSPF will still take priority. We need to use wide metrics, which is required for v6 address family support in any case. And note that we only want to use Level-2 IS. Cisco IOS default is L1L2. As for v6 addressing, the network operator has a choice of using global unicast addresses, or link local addressing with unnumbered interfaces. The slide shows a configuration example using Cisco IOS. We have a loopback interface with v4 and v6 addresses on it, and we have an Ethernet interface, using a v4 global address, and a v6 unnumbered address based on the loopback. We have configured IS-IS in the v4 address family, and we configured IS-IS in the v6 address family on that interface, and set metrics for both v4 and v6. The following slide, which is the second part of the configuration, shows another example interface and how the IS-IS process itself is configured. Note the key features. The 'net' address, the passive interface for the loopback, we've set the IS type to be level-2 only, we've set wide metrics, and we've configured the overload bit, so that IS-IS will not announce it being best path until BGP is operational. For the v6 address family we have created multi-topology, so we're running a separate topology, and we've also set the overload bit for the v6 configuration. I have a similar configuration on the screen now, for IOS XR. The same interfaces and then the same IS-IS configuration. Notice that IOS XR, the IS-IS configuration for each interface is configured within the IS-IS process itself. Other vendors use this style of configuration syntax also.

© Produced by Philip Smith and the Network Startup Resource Center, through the University of Oregon.

Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)
This is a human-readable summary of (and not a substitute for) the license. Disclaimer. You are free to: Share — copy and redistribute the material in any medium or format Adapt — remix, transform, and build upon the material The licensor cannot revoke these freedoms as long as you follow the license terms. Under the following terms: Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. NonCommercial — You may not use the material for commercial purposes. No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.