# OSPFv3

> Source: Courseiva IT Certification Glossary — https://courseiva.com/glossary/ospfv3

## Quick definition

OSPFv3 is a routing protocol used by routers to share information about how to reach IPv6 destinations. It helps routers find the best paths for IPv6 traffic across a network. Unlike OSPFv2, which works with IPv4, OSPFv3 is designed specifically for IPv6. It still uses the same basic algorithm to calculate shortest paths, but it has been updated to handle the larger IPv6 addresses and new features.

## Simple meaning

Think of OSPFv3 as a GPS navigation system for data traveling across the internet or a large company network, but only for vehicles that speak IPv6. Imagine you have a fleet of delivery trucks that use a new kind of address book called IPv6. Each truck needs to know the fastest route to deliver packages. OSPFv3 is like a team of coordinators at each hub (router) who constantly talk to each other about which roads are open, which are closed, and how long each road takes. They collaborate to build a master map so that every truck takes the most efficient path.

In everyday terms, if you were using a car’s GPS, the GPS would receive live traffic updates, road closures, and alternative routes. OSPFv3 does the same thing for routers. It sends updates whenever a link goes down or a new network segment appears. The routers then recalculate the best route. The big difference from its predecessor, OSPFv2, is that OSPFv3 works purely with IPv6 addresses, which are much longer and more numerous than IPv4 addresses. It also separates the routing information from the actual IP addresses, making it more flexible.

For a learner, understanding OSPFv3 means knowing that it is a link-state protocol. Each router has a complete picture of the network topology, not just a list of next hops. This is like each coordinator having a full map of the city, rather than just knowing the next intersection. This allows for faster convergence when something changes, because every router can independently calculate the new best path using the same map.

## Technical definition

OSPFv3, defined in RFC 5340, is a link-state routing protocol for IPv6. It is an evolution of OSPFv2, which routes IPv4 traffic. The core algorithm remains Dijkstra's Shortest Path First (SPF) algorithm, used to compute the shortest path tree to all destinations. However, OSPFv3 introduces several fundamental changes to accommodate IPv6 and improve protocol design.

One of the most significant changes is that OSPFv3 runs per-link rather than per-subnet. In OSPFv2, each subnet (IP network) was associated with a specific interface. In OSPFv3, multiple IPv6 prefixes can be associated with a single link, and the protocol itself does not carry IPv4 or IPv6 addresses in its router links or network LSAs. Instead, addresses are carried in separate LSAs: Intra-Area-Prefix-LSAs and Link-LSAs. This decoupling allows OSPFv3 to support multiple IPv6 prefixes on the same link without needing to change the core LSA types.

Another critical change is the use of IPv6 link-local addresses for neighbor discovery and next-hop information. OSPFv3 routers communicate using their link-local addresses, which are automatically configured on each interface. This simplifies configuration because the global or unique local IPv6 addresses do not need to be explicitly specified for OSPFv3 adjacencies.

OSPFv3 also introduces new LSA types. The Link-LSA (type 8) carries the link-local address and list of IPv6 prefixes associated with a link. The Intra-Area-Prefix-LSA (type 9) associates prefixes with either a router or a network (transit link). These replace the functionality of the type 1 and type 2 LSAs in OSPFv2 for prefix information. The Inter-Area-Prefix-LSA (type 3) and Inter-Area-Router-LSA (type 4) work similarly to their OSPFv2 counterparts but carry IPv6 prefixes.

Authentication in OSPFv3 is handled differently. OSPFv2 used simple or MD5 authentication within the packet header. OSPFv3 relies on IPsec (Authentication Header or Encapsulating Security Payload) for integrity and authentication. This provides stronger security and aligns with the IPv6 security framework. The OSPFv3 packet header contains no authentication fields.

In real IT implementations, OSPFv3 is configured on routers in enterprise and service provider networks that have migrated to IPv6. It supports the same area structure, with a backbone area (area 0) and standard areas. It also supports virtual links, stub areas, and not-so-stubby-areas (NSSA) through the OSPFv3 NSSA LSA (type 7), though the implementation details differ slightly from OSPFv2. Multi-area adjacency is also supported in some implementations, allowing routers to form adjacencies over a shared link for multiple areas.

## Real-life example

Imagine you are the logistics manager for a large delivery company that has just switched to a new, more detailed address system called HyperAddress (our IPv6). Your delivery hubs (routers) are spread across the city, and each hub needs to know the best way to send packages to any address. The old address system, SimpleAddress (IPv4), was running out of numbers, so the company moved to HyperAddress, which has an almost unlimited number of addresses.

To manage deliveries efficiently, you install a new communication system called Coordinated Hub Mapping (OSPFv3). In this system, each hub continuously talks to its neighboring hubs using a dedicated intercom channel that uses only the hub's internal, local radio frequency (this is like the link-local address in IPv6). They do not need to use their full HyperAddress for these conversations, which saves time and reduces complexity.

Every time a road is closed for construction or a new warehouse is added, the hub that notices the change sends out a quick bulletin to all other hubs. Each hub then independently updates its master map of the city, using a consistent formula to calculate the shortest path to every HyperAddress. Because all hubs have the same map, they all come to the same conclusion about the best route.

This system is much more efficient than the old one because it separates the map of the roads from the list of addresses. In the old system, each address was tied to a specific road segment. In the new system, the map shows the roads, and a separate list shows which addresses are on which road. This means if you add a new street with hundreds of HyperAddresses, you only need to update the address list, not the entire road map. This is how OSPFv3 uses different LSA types for topology and prefix information.

## Why it matters

OSPFv3 matters because IPv6 deployment is growing globally, driven by the exhaustion of IPv4 addresses. As organizations enable IPv6 on their networks, they need a robust, scalable interior gateway protocol (IGP) to route that traffic. OSPFv3 is the natural choice for networks already running OSPF for IPv4, as it allows a smooth transition and uses the same operational concepts. Network engineers and architects must understand OSPFv3 to design and maintain dual-stack or IPv6-only networks.

From a practical IT standpoint, OSPFv3 offers improved security by leveraging IPsec, which is mandatory in some government and financial sector networks. Its separation of topology and prefix information makes it more flexible and easier to manage in environments with complex addressing schemes. For example, a service provider that assigns multiple IPv6 prefixes to a single customer link can do so without reconfiguring the OSPFv3 adjacency.

For certification candidates, OSPFv3 is a core topic because it demonstrates a deep understanding of how routing protocols adapt to changes in the network layer. It is not just a minor update; it is a fundamental redesign that reflects best practices in protocol design. Employers value professionals who can troubleshoot OSPFv3 issues, such as neighbor adjacency problems caused by mismatched MTU or incorrect IPsec configuration. Without this knowledge, a network engineer might mistakenly apply OSPFv2 concepts to an OSPFv3 problem, leading to configuration errors and network outages.

OSPFv3 supports multiple instances per link, which is useful in MPLS or VPN environments. This capability, called OSPFv3 instance ID, allows multiple virtual routing tables to share the same physical link. Understanding this is critical for advanced networking roles. OSPFv3 is not just a protocol for the future; it is a protocol for the present in any network that has started the IPv6 transition.

## Why it matters in exams

OSPFv3 appears in several major IT certification exams, particularly those from Cisco, Juniper, and CompTIA. For Cisco, it is a key topic in the CCNP Enterprise (350-401 ENCOR) and CCIE Enterprise Infrastructure exams. In the ENCOR exam blueprint, OSPFv3 is listed under Layer 3 technologies, and candidates must be able to configure and verify OSPFv3, including IPv6 addressing, neighbor adjacencies, and routing table verification. The exam may present troubleshooting scenarios where OSPFv3 adjacencies fail due to mismatched area IDs, MTU, or authentication issues. Multiple-choice questions often test the differences between OSPFv2 and OSPFv3, such as the types of LSAs, the use of link-local addresses, and the authentication mechanism.

For Juniper JNCIP-SP or JNCIP-ENT, OSPFv3 configuration and operation are also tested. Juniper uses a different configuration syntax but the same underlying principles. Candidates need to understand how OSPFv3 interacts with other IPv6 routing protocols, such as static routes or BGP. The exam might ask about the purpose of Link-LSAs or Intra-Area-Prefix-LSAs, or require analysis of an OSPFv3 database output.

CompTIA Network+ and Security+ cover OSPFv3 at a more conceptual level. Network+ expects candidates to know that OSPFv3 is the IPv6 version of OSPF and that it uses the SPF algorithm. They may not need to configure it, but they must understand its role in a routed network. Security+ touches on OSPFv3 in the context of IPsec for authentication, as part of secure routing.

Exam question types include: multiple-choice questions asking which LSA type carries IPv6 prefixes in OSPFv3, or which address type is used for neighbor discovery. Simulation questions may ask you to configure OSPFv3 on a router, set the router ID, and enable IPv6 routing. Troubleshooting questions might show a partial configuration with a missing network command (though OSPFv3 uses the network command differently) or an incorrect passive interface setting. Performance-based questions might require you to interpret the output of 'show ipv6 ospf neighbor' or 'show ipv6 ospf database'.

In the CCIE lab exam, OSPFv3 is often combined with other technologies like MPLS, VPNv6, or multicast. Candidates must be able to configure OSPFv3 over a GRE tunnel, implement virtual links in IPv6, and verify multi-area adjacency. The exam expects a deep understanding of the OSPFv3 packet structure, LSA flooding, and the impact of timers on convergence. Overall, OSPFv3 is not just a checkbox topic; it is a foundational protocol that examiners use to test a candidate's understanding of IPv6 and routing protocol design.

## How it appears in exam questions

Questions about OSPFv3 in certification exams come in several common patterns. The first is scenario-based questions that describe a network with multiple routers running OSPFv3. For example, a question might say: 'Router A and Router B are OSPFv3 neighbors, but Router A cannot reach a network behind Router B. Given the following partial configurations and debug output, what is the most likely cause?' The answer often relates to a mismatched area ID, a passive interface blocking hello packets, or an MTU mismatch that prevents adjacency formation. Candidates must be able to deduce the issue from the provided outputs.

Another common pattern is configuration questions. These might ask: 'Which command enables OSPFv3 on an interface on a Cisco router?' The correct answer is 'ipv6 ospf 1 area 0' under interface configuration mode. However, some questions might try to confuse you with 'router ospf 1' (which is for IPv4) or 'ipv6 router eigrp 100'. Knowing the exact syntax is critical. Similarly, questions may ask for the command to set the router ID for OSPFv3, which is 'router-id 1.1.1.1' under 'ipv6 router ospf 1'.

Troubleshooting questions often show the output of 'show ipv6 ospf neighbor' and ask why a neighbor is stuck in the INIT or EXSTART state. Common reasons include a mismatched dead interval, a missing IPv6 link-local address, or a firewall blocking OSPFv3 packets (protocol 89 for IPv6). Another trick is a question about OSPFv3 authentication: since OSPFv3 uses IPsec, a candidate might mistakenly answer that it uses MD5 or simple authentication. The correct answer is IPsec, but the question might be phrased as 'Which security mechanism does OSPFv3 use by default?'

Multiple-choice questions often test the differences between OSPFv2 and OSPFv3. For instance, 'Which LSA type in OSPFv3 carries IPv6 prefixes?' The answer is Intra-Area-Prefix-LSA (type 9) or Inter-Area-Prefix-LSA (type 3), depending on the scope. Another frequent question: 'Which address type do OSPFv3 routers use to form adjacencies?' The answer is link-local IPv6 addresses.

Finally, there are design questions: 'A network administrator is migrating from OSPFv2 to OSPFv3. Which two features must be considered?' The answers might include the removal of authentication from the packet header and the introduction of Link-LSAs. These questions require a conceptual understanding, not just memorization of commands. To succeed, candidates should practice interpreting OSPFv3 show commands and debug outputs, and be able to map the theory to real-world scenarios.

## Example scenario

You are a network administrator for a university that has just been allocated a large IPv6 address block. The university has three campus routers: RouterA, RouterB, and RouterC, all connected in a triangle. You need to configure OSPFv3 so that all IPv6 networks can communicate.

First, you enable IPv6 routing globally on all routers with the command 'ipv6 unicast-routing'. Then, you configure OSPFv3. On RouterA, you enter 'ipv6 router ospf 1' and assign a router ID of '1.1.1.1'. You enter interface configuration mode for the interface connecting to RouterB and type 'ipv6 ospf 1 area 0'. You do the same for the interface to RouterC. On RouterB and RouterC, you perform similar configurations, ensuring all interfaces are in area 0. You also advertise the IPv6 networks connected to each router's LAN segments by enabling OSPFv3 on those interfaces as well.

After configuring, you check the OSPFv3 neighbors using 'show ipv6 ospf neighbor'. You see that RouterA sees RouterB in the FULL state, but RouterC is stuck in EXSTART. You suspect an MTU issue. You check the MTU on the link between RouterA and RouterC and find that RouterA has an MTU of 1500 while RouterC has 1492 (due to a VPN tunnel). You fix the MTU mismatch by setting both interfaces to 1492. The adjacency then comes up to FULL.

Now you verify the routing table with 'show ipv6 route ospf'. You see routes to the IPv6 networks on RouterB and RouterC. You test connectivity by pinging an IPv6 address on RouterB's LAN from a host on RouterA. The ping succeeds. This scenario demonstrates how OSPFv3 dynamically learns IPv6 routes and how a common issue (MTU mismatch) can disrupt the adjacency. It also shows the importance of verifying neighbor states during troubleshooting.

## Common mistakes

- **Mistake:** Using OSPFv2 commands for OSPFv3, like typing 'network' under the OSPF process to advertise networks.
  - Why it is wrong: OSPFv3 does not use the 'network' command under the router process. Instead, OSPFv3 is enabled directly on the interface with 'ipv6 ospf <process-id> area <area-id>'. Trying to use OSPFv2 style network statements will not work.
  - Fix: Remove any 'network' statements from the OSPFv3 process and enable OSPFv3 on each interface individually using the interface-level command.
- **Mistake:** Forgetting to enable IPv6 unicast routing globally before configuring OSPFv3.
  - Why it is wrong: OSPFv3 requires IPv6 routing to be enabled on the router. Without 'ipv6 unicast-routing', the router will not process IPv6 packets, and OSPFv3 hellos will not be sent or received.
  - Fix: Enter global configuration mode and issue the command 'ipv6 unicast-routing'. Then verify with 'show ipv6 interface brief' to ensure interfaces have IPv6 addresses.
- **Mistake:** Assuming OSPFv3 uses the same authentication method as OSPFv2 (MD5 or simple).
  - Why it is wrong: OSPFv3 does not have authentication fields in the packet header. It relies on IPsec for authentication and encryption. Configuring 'ip ospf authentication' will have no effect on OSPFv3.
  - Fix: Use IPsec policies to secure OSPFv3 traffic if needed. For basic exams, simply know that OSPFv3 uses IPsec, not standalone authentication commands.
- **Mistake:** Confusing OSPFv3 router ID with IPv6 addresses. Setting the router ID to an IPv6 address.
  - Why it is wrong: The OSPFv3 router ID is still a 32-bit dotted decimal format, just like in OSPFv2. It is not an IPv6 address. For example, 'router-id 1.1.1.1' is correct.
  - Fix: Always set the router ID using a 32-bit value in dotted decimal notation. Use an IP address from your IPv4 loopback or any unused number, but ensure it is unique across the OSPF domain.

## Exam trap

{"trap":"The exam presents a scenario where OSPFv3 adjacencies are not forming, and gives the option to configure 'ipv6 ospf authentication md5 key-chain MYKEY' as a fix.","why_learners_choose_it":"Learners are familiar with OSPFv2 authentication and assume OSPFv3 works the same way. They see the word 'authentication' and think it is the correct solution to an adjacency problem, not realizing OSPFv3 does not support that command.","how_to_avoid_it":"Remember that OSPFv3 authentication is handled by IPsec, not by OSPFv3 itself. If the question is about OSPFv3 adjacency issues, look for other causes like mismatched area IDs, MTU, or passive interfaces. Do not apply OSPFv2 authentication logic to OSPFv3."}

## Commonly confused with

- **OSPFv3 vs OSPFv2:** OSPFv2 is the version for IPv4, while OSPFv3 is for IPv6. OSPFv2 uses network statements under the router process, while OSPFv3 is configured per interface. OSPFv2 includes authentication in the packet header; OSPFv3 relies on IPsec. OSPFv2 LSAs carry IPv4 addresses directly; OSPFv3 uses separate LSAs (Link-LSA and Intra-Area-Prefix-LSA) for prefix information. (Example: If you see a configuration with 'router ospf 1' and 'network 10.0.0.0 0.255.255.255 area 0', that is OSPFv2. If you see 'ipv6 router ospf 1' and 'ipv6 ospf 1 area 0' under the interface, that is OSPFv3.)
- **OSPFv3 vs EIGRP for IPv6:** EIGRP for IPv6 is also a routing protocol for IPv6, but it is a distance-vector protocol (with some advanced features), while OSPFv3 is a link-state protocol. EIGRP uses Diffusing Update Algorithm (DUAL), whereas OSPFv3 uses the SPF algorithm. EIGRP configuration for IPv6 uses a named mode or 'ipv6 router eigrp <as-number>', while OSPFv3 uses 'ipv6 router ospf <process-id>'. (Example: In a Cisco environment, you might configure 'ipv6 router eigrp 100' vs 'ipv6 router ospf 1'. The decision between them often depends on network size and vendor preference.)
- **OSPFv3 vs RIPng:** RIPng (RIP Next Generation) is a simple distance-vector routing protocol for IPv6. It uses hop count as a metric (max 15), while OSPFv3 uses cost based on bandwidth. RIPng converges much slower than OSPFv3 and is not suitable for large networks. OSPFv3 is more scalable and has faster convergence due to its link-state nature. (Example: RIPng is sometimes used in small or academic networks for simplicity. OSPFv3 is used in enterprise and service provider networks where convergence time and scalability are critical.)

## Step-by-step breakdown

1. **Enable IPv6 Routing** — Before OSPFv3 can work, the router must be able to process IPv6 packets. This is done globally with 'ipv6 unicast-routing'. Without this, IPv6 interfaces may have addresses but the router will not forward IPv6 traffic or participate in OSPFv3.
2. **Configure OSPFv3 Process** — Create the OSPFv3 routing process using 'ipv6 router ospf <process-id>'. The process ID is locally meaningful and does not need to match between routers. A router ID must be assigned in dotted decimal format, for example, 'router-id 1.1.1.1'.
3. **Enable OSPFv3 on Interfaces** — On each interface that should participate in OSPFv3, enter interface configuration mode and issue 'ipv6 ospf <process-id> area <area-id>'. This step replaces the network statement used in OSPFv2. The interface must have an IPv6 address configured, typically a link-local address and optionally a global address.
4. **Verify Neighbor Adjacencies** — Use 'show ipv6 ospf neighbor' to see if the router has formed adjacencies with its neighbors. Expected states are FULL for DR/BDR relationships or FULL/DR for others. If neighbors are stuck in other states (INIT, EXSTART), there is a problem that needs troubleshooting.
5. **Check the OSPFv3 Database** — Use 'show ipv6 ospf database' to view the LSAs collected from neighbors. This includes Router LSAs, Network LSAs, Link-LSAs, and Intra-Area-Prefix-LSAs. Ensuring that the database contains the expected LSAs confirms that OSPFv3 is exchanging routing information correctly.
6. **Verify the IPv6 Routing Table** — Use 'show ipv6 route ospf' to see the OSPFv3-learned routes. The routes will be marked with 'O' for intra-area, 'OI' for inter-area, or 'ON' for NSSA. If expected routes are missing, check for passive interfaces, area mismatches, or filters.

## Practical mini-lesson

OSPFv3 is not a complete rewrite of OSPFv2, but it is a careful adaptation that leverages IPv6 features while retaining the proven SPF algorithm. For a network professional, the most important practical difference is the configuration model. In OSPFv2, you define a network statement under the router process that matches an interface IP. In OSPFv3, you enable the protocol directly on the interface. This shift demands a different mindset when designing and troubleshooting networks.

When you enable OSPFv3 on an interface, the router automatically uses its IPv6 link-local address as the source for Hello packets and as the next hop in routing updates. This is a significant change because in OSPFv2, the router uses the interface's IPv4 address for these purposes. One consequence is that you must ensure that IPv6 link-local addresses are unique on each link, which they generally are by default. However, if you manually configure link-local addresses, you must avoid duplication.

Another practical consideration is the handling of multiple IPv6 prefixes on a single interface. In OSPFv2, if you wanted to advertise multiple subnets on a single interface, you needed multiple network statements. In OSPFv3, you simply configure all IPv6 addresses on the interface, and OSPFv3 advertises them via Intra-Area-Prefix-LSAs. This simplifies configuration in environments with complex addressing, such as when using multiple VLANs or prefix delegation.

What can go wrong? One common issue is the missing 'ipv6 unicast-routing' command. If you forget this, the router will have IPv6 addresses but will not route or run OSPFv3. Another issue is the MTU mismatch, which can cause adjacencies to remain stuck in EXSTART state. This is because OSPFv3 packets may be fragmented or dropped if the MTU differs between two ends. The solution is to ensure consistent MTU on both sides of a link. Also, remember that OSPFv3 uses multicast addresses FF02::5 (all OSPF routers) and FF02::6 (DR/BDR). If these multicast addresses are blocked by switch configurations, adjacencies will not form.

Security is another practical area. In many enterprise networks, IPsec is required for OSPFv3 authentication. Configuring IPsec for OSPFv3 can be complex, but it is essential for compliance. Without IPsec, OSPFv3 traffic is vulnerable to spoofing and replay attacks. Some network engineers mistakenly think that OSPFv3 is inherently secure because it uses IPv6, but this is not true. You must still implement security measures.

Finally, understand the OSPFv3 database. The 'show ipv6 ospf database' command displays LSAs with different types. A type 0x2001 is a Router LSA, 0x2002 is a Network LSA, 0x2008 is a Link LSA, and 0x2009 is an Intra-Area-Prefix LSA. Being able to read this output helps in diagnosing routing issues. For example, if you see a missing Intra-Area-Prefix LSA for a network you expect, check that the network is configured on the correct interface and that the interface is not passive. Mastering OSPFv3 in practice requires hands-on configuration, careful troubleshooting, and a solid grasp of how it differs from OSPFv2.

## Memory tip

Remember OSPFv3 as 'OSPF for IPv6, no network statements, use IPsec for security'.

## FAQ

**Can I use the same OSPF process for both OSPFv2 and OSPFv3?**

No, OSPFv2 and OSPFv3 have separate routing processes. You must configure one for IPv4 and one for IPv6. They operate independently.

**Does OSPFv3 require a router ID?**

Yes, OSPFv3 requires a 32-bit router ID in dotted decimal format, just like OSPFv2. It can be set manually or will default to the highest IPv4 address on a loopback interface.

**Is OSPFv3 authentication optional?**

Authentication is not built into OSPFv3 itself, but IPsec can be used to secure OSPFv3 traffic. In some environments it is required, but it is not mandatory by default.

**What are the multicast addresses used by OSPFv3?**

OSPFv3 uses FF02::5 for all OSPF routers and FF02::6 for the designated router and backup designated router.

**How does OSPFv3 handle multiple IPv6 prefixes on a single interface?**

OSPFv3 uses Link-LSAs (type 8) and Intra-Area-Prefix-LSAs (type 9) to advertise multiple prefixes associated with a link, without requiring multiple network statements.

**Can OSPFv3 and OSPFv2 run on the same router simultaneously?**

Yes, a router can run both OSPFv2 for IPv4 and OSPFv3 for IPv6 at the same time. They do not interfere with each other.

## Summary

OSPFv3 is the IPv6-optimized version of the OSPF routing protocol, designed to handle the larger address space and new features of IPv6 while retaining the core link-state algorithm. It differs from OSPFv2 in key ways: configuration is per-interface instead of using network statements, it uses IPv6 link-local addresses for neighbor discovery, and it relies on IPsec for authentication. The protocol introduces new LSA types, such as Link-LSA and Intra-Area-Prefix-LSA, to separate topology from prefix information. This makes OSPFv3 more flexible and scalable for modern networks.

Why does this matter? As IPv6 adoption accelerates, network professionals must be proficient in OSPFv3 to design, configure, and troubleshoot IPv6 networks. Certification exams from Cisco, Juniper, and CompTIA test OSPFv3 knowledge in both conceptual and practical forms. Candidates must understand the differences between OSPFv2 and OSPFv3, be able to configure it correctly, and diagnose common issues like MTU mismatches or missing IPv6 unicast routing.

The exam takeaway is clear: OSPFv3 is not just a minor update; it is a significant protocol that requires specific study. Focus on the configuration syntax, LSA types, and the absence of OSPFv2-style authentication. Practice with simulators or lab equipment to solidify your understanding. By mastering OSPFv3, you demonstrate a readiness to handle the IPv6 networks of today and tomorrow.

---

Practice questions and the full interactive page: https://courseiva.com/glossary/ospfv3