LSP Generation

Understanding LSP Generation

Link State PDUs (LSPs) are the heart of IS-IS routing. Each router generates LSPs to describe its local topology, which are then flooded throughout the network to build a complete view of the network topology.

LSP Generation Triggers

Routers generate new LSPs or update existing ones when specific network events occur. Understanding these triggers is crucial for network stability and convergence.

Immediate Triggers
  • Interface state change (up/down)
  • New adjacency formation
  • Adjacency loss
  • Metric change on interface
  • IP address configuration change
Periodic Triggers
  • LSP refresh timer (default 900 seconds)
  • Checksum validation failure
  • LSP aging (approaching expiration)
  • Manual LSP regeneration
  • Configuration changes requiring update

LSP Generation Process

1
Trigger Event

Network change occurs requiring LSP update

→
2
Information Collection

Gather current interface and adjacency data

→
3
LSP Construction

Build LSP with header and TLVs

→
4
Sequence Number

Increment sequence number for newer version

→
5
Flooding

Flood new LSP to all neighbors

LSP Content

LSP Content Example

# Router LSP Content
LSP ID: 1921.6800.0001.00-00
Sequence: 0x00000005
Remaining Lifetime: 1199 seconds

TLVs:
- Area Addresses (TLV 1): 49.0001
- IS Neighbors (TLV 2): 
  * 1921.6800.0002 (metric: 10)
  * 1921.6800.0003 (metric: 10)
- IP Interface Address (TLV 132): 192.168.1.1
- IP Internal Reachability (TLV 128):
  * 192.168.1.0/24 (metric: 0)
  * 10.1.1.0/30 (metric: 10)

LSP Flooding

Flooding Mechanism

LSP flooding is the process by which LSPs are distributed throughout the IS-IS domain. The flooding mechanism ensures all routers receive a copy of every LSP in their level.

Flooding Algorithm

1. LSP Reception

Router receives LSP from neighbor

2. Database Check

Compare with existing LSP in database

3. Newer LSP Decision

Determine if received LSP is newer

4. Database Update

Install newer LSP in database

5. Flood to Neighbors

Forward LSP to all other neighbors

6. Acknowledge

Send acknowledgment to originating neighbor

LSP Comparison Criteria

Comparison Received LSP Database LSP Action
Sequence Number Higher Lower Install and flood
Sequence Number Lower Higher Send newer LSP back
Sequence Number Equal Equal Compare checksum
Lifetime 0 (purge) Non-zero Accept purge and flood

Flooding on Different Network Types

Broadcast Networks
Flooding Process:
  1. LSP sent to all routers multicast address
  2. DIS coordinates flooding
  3. CSNP used for synchronization
  4. PSNP used for acknowledgments
Point-to-Point
Flooding Process:
  1. LSP sent directly to neighbor
  2. No DIS coordination needed
  3. Direct acknowledgment required
  4. PSNP acknowledgment mandatory

Flooding Optimization

Flooding Optimizations

  • Split Horizon: Don't flood LSP back to sender
  • Duplicate Detection: Avoid processing same LSP multiple times
  • Rate Limiting: Control LSP generation frequency
  • Mesh Groups: Reduce flooding in full-mesh topologies

SPF Calculation

Dijkstra's Shortest Path First Algorithm

IS-IS uses Dijkstra's SPF algorithm to compute shortest paths to all destinations. The algorithm operates on the link state database to build a shortest path tree rooted at the calculating router.

SPF Algorithm Overview

1
Initialize

Set distance to self as 0, all others as infinity

2
Add to TENT

Add self to tentative list with distance 0

3
Select Node

Pick node with smallest distance from TENT

4
Move to PATHS

Move selected node to shortest paths tree

5
Update Neighbors

Update distances to neighbors of selected node

6
Repeat

Continue until TENT list is empty

SPF Example

Network Topology

    A (metric 10) B (metric 15) C
     \             |             /
      \            |            /
       \(metric 20)|           /(metric 25)
        \          |          /
         \         |         /
          \        |        /
           \   (metric 5)  /
            \      |      /
             \     |     /
              \    |    /
               \   |   /
                \  |  /
                 \ | /
                   D
SPF Calculation from Router A:
Step TENT List PATHS List Action
1 A(0) - Initialize with A
2 B(10), D(20) A(0) Add A's neighbors
3 D(20), C(25) A(0), B(10) Select B, add C
4 C(25) A(0), B(10), D(20) Select D (better via A-D than A-B-D)
5 - A(0), B(10), D(20), C(25) Select C, TENT empty

SPF Triggers

Full SPF
  • New LSP received
  • LSP with topology change
  • LSP aged out or purged
  • Significant metric change
Partial SPF
  • Leaf prefix changes only
  • No topology changes
  • Route add/delete operations
  • Less CPU intensive

IS-IS Metrics

Metric Types

IS-IS supports multiple metric systems to accommodate different network requirements and provide flexibility in path selection.

Narrow Metrics
  • Range: 1-63 per link
  • Path Cost: 1-1023 total
  • Default: 10 for all interfaces
  • Usage: Legacy deployments
  • TLV: 128, 130 (IP Reachability)
Wide Metrics
  • Range: 1-16,777,215 per link
  • Path Cost: Up to 4,261,412,864
  • Scalability: Much better for large networks
  • Usage: Modern deployments
  • TLV: 135, 236 (Extended IP Reachability)

Metric Configuration

IS-IS Metric Configuration Examples

# Enable wide metrics globally
router isis
 metric-style wide

# Configure interface metrics
interface GigabitEthernet0/0
 isis metric 100
 isis metric 150 level-2

# Auto-calculate metrics based on bandwidth
router isis
 metric-style wide
 auto-cost reference-bandwidth 100000

# Manual metric assignment
interface Serial0/0
 isis metric 1000 level-1
 isis metric 1500 level-2

Metric Calculation Methods

Method Calculation Use Case Advantage
Fixed Manual assignment Simple networks Predictable behavior
Bandwidth-based Reference BW / Interface BW Mixed speed links Reflects link capacity
Delay-based Based on link delay Latency-sensitive apps Optimizes for delay
Administrative Policy-based assignment Traffic engineering Manual control

Metric Best Practices

Metric Design Guidelines

  • Use Wide Metrics: Always use wide metrics for new deployments
  • Consistent Methodology: Use same metric calculation method network-wide
  • Room for Growth: Leave headroom in metric values for future adjustments
  • Document Strategy: Clearly document metric assignment strategy
  • Test Changes: Verify routing changes before implementing metric modifications

Route Leaking

Understanding Route Leaking

Route leaking is the process of redistributing routes between IS-IS levels. By default, IS-IS follows strict hierarchical routing, but route leaking allows for more flexible routing policies.

Default IS-IS Routing Behavior

Level-1 Routing

Routes to destinations within the same area. Uses default route to Level-1-2 router for inter-area traffic.

Level-2 Routing

Routes between areas and to external destinations. Maintains area-level reachability information.

Level-1-2 Routing

Participates in both levels. Sets attached bit in Level-1 LSPs to indicate external connectivity.

Route Leaking Scenarios

Level-2 to Level-1
When to Use:
  • Provide specific inter-area routes to L1 routers
  • Avoid suboptimal routing through default
  • Implement traffic engineering policies
Configuration:
redistribute isis level-2 into level-1
Level-1 to Level-2
When to Use:
  • Advertise specific intra-area prefixes
  • Provide reachability to stub networks
  • Support multi-homed area designs
Configuration:
redistribute isis level-1 into level-2

Route Leaking Configuration

Route Leaking Examples

# Basic route leaking
router isis
 redistribute isis level-2 into level-1 distribute-list 10
 redistribute isis level-1 into level-2 distribute-list 20

# Access lists to control leaked routes
access-list 10 permit 172.16.0.0 0.0.255.255
access-list 10 deny any

access-list 20 permit 192.168.0.0 0.0.255.255
access-list 20 deny any

# Route map based leaking
router isis
 redistribute isis level-2 into level-1 route-map L2_TO_L1

route-map L2_TO_L1 permit 10
 match ip address prefix-list IMPORTANT_PREFIXES
 set metric 100

Route Leaking Considerations

Important Considerations

  • Routing Loops: Carefully control which routes are leaked to prevent loops
  • Database Size: Route leaking increases LSDB size, especially in Level-1
  • Convergence: More routes may impact convergence time
  • Memory Usage: Additional routes consume more router memory
  • Filtering: Always use filtering to control leaked routes

Advanced Route Leaking

Selective Route Leaking with Route Maps

# Advanced route leaking with metric manipulation
route-map LEAK_L2_TO_L1 permit 10
 match ip address prefix-list CRITICAL_SERVERS
 set metric 50
 set metric-type internal

route-map LEAK_L2_TO_L1 permit 20
 match ip address prefix-list BACKUP_PATHS
 set metric 200
 set metric-type internal

route-map LEAK_L2_TO_L1 deny 30

# Apply the route map
router isis
 redistribute isis level-2 into level-1 route-map LEAK_L2_TO_L1
Previous: Chapter 3 - PDUs and Packets Next: Chapter 5 - IS-IS Configuration