Juniper Updated JN0-664 Exam Questions and Answers by ruqayyah

Intermediate System to Intermediate System (IS-IS) is a link-state routing protocol used to move information efficiently within a computer network. It uses a series of Protocol Data Units (PDUs) to manage the network's topology and ensure consistency across all routers in the network. Specifically, Link State PDUs (LSPs), Complete Sequence Number PDUs (CSNPs), and Partial Sequence Number PDUs (PSNPs) play crucial roles in this process.

1. **PSNPs (Partial Sequence Number PDUs)**:

- **Acknowledge a received LSP**: PSNPs are used to acknowledge the receipt of LSPs. When a router receives an LSP, it sends a PSNP back to the sender to confirm that the LSP has been received.

- **Request a missing LSP**: PSNPs are also used to request missing LSPs. If a router identifies a missing LSP based on sequence numbers, it can send a PSNP to request the specific LSP from its neighbors.

2. **CSNPs (Complete Sequence Number PDUs)**:

- **Summarize LSPs**: CSNPs are used to summarize all the LSPs known to a router. They are typically sent at regular intervals to provide a complete list of LSPs in a database. They are not used to acknowledge or request specific LSPs but provide an overview of all LSPs for database synchronization.

Based on this understanding, let's evaluate the statements:

- **A. PSNPs are used to acknowledge a received LSP.**

- Correct. PSNPs serve the purpose of acknowledging LSPs received from other routers.

- **B. CSNPs are used to acknowledge a received LSP.**

- Incorrect. CSNPs are not used for acknowledging LSPs; they are used to provide a summary of all LSPs.

- **C. CSNPs are used to request a missing LSP.**

- Incorrect. CSNPs are not used to request missing LSPs; this is the role of PSNPs.

- **D. PSNPs are used to request a missing LSP.**

- Correct. PSNPs are used to request specific missing LSPs when a router detects that it is missing information.

**Conclusion**:

The correct statements about IS-IS are:

**A. PSNPs are used to acknowledge a received LSP.**

**D. PSNPs are used to request a missing LSP.**

**References**:

- Juniper Networks Documentation on IS-IS: [IS-IS Overview](https://www.juniper.net/documentation/en_US/junos/topics/concept/is-is-routing-overview.html)

- RFC 1195, Use of OSI IS-IS for Routing in TCP/IP and Dual Environments: [RFC 1195](https://tools.ietf.org/html/rfc1195) which details the operation and use of IS-IS, including the roles of PSNPs and CSNPs.

as by default, the MED attribute is only compared for routes received from the same neighbouring AS. The next feasible tiebreaker in the BGP route selection algorithm would be Router ID.

TLVs are tuples of (Type, Length, Value) that can be advertised in IS-IS packets. TLVs can carry different kinds of information in the Link State Packets (LSPs). IS-IS supports both narrow and wide metrics for link costs. Narrow metrics use a single octet to encode the link cost, while wide metrics use three octets. Narrow metrics have a maximum value of 63, while wide metrics have a maximum value of 16777215. If there are both narrow and wide metric devices in the topology, IS-IS will advertise two TLVs for each link: one with the narrow metric and one with the wide metric. This allows backward compatibility with older devices that only support narrow metrics12.

IS-IS is a link-state routing protocol that uses the Dijkstra algorithm to compute the shortest paths between nodes in a network. The Dijkstra algorithm maintains three data structures: a tree database, a candidate database, and a link-state database (LSDB). The tree database contains the nodes that have been visited and their shortest distances from the source node. The candidate database contains the nodes that have not been visited yet and their tentative distances from the source node. The LSDB contains the topology information of the network, such as the links and their costs.

The Dijkstra algorithm works as follows:

The local router moves its own local tuples into the tree database. A tuple consists of a node ID, a distance, and a parent node ID. The local router’s tuple has a distance of zero and no parent node.

The local router moves its neighbors’ tuples into the candidate database. The neighbors’ tuples have distances equal to the costs of the links to them and parent node IDs equal to the local router’s node ID.

The local router selects the tuple with the lowest distance from the candidate database and moves it to the tree database. This tuple becomes the current node.

The local router updates the distances of the current node’s neighbors in the candidate database by adding the current node’s distance to the link costs. If a shorter distance is found, the parent node ID is also updated.

The algorithm repeats steps 3 and 4 until either the destination node is reached or the candidate database is empty.