10.1 IS-IS Encapsulation and NET Identity
Key Takeaways
IS-IS uses LLC/ISO encapsulation rather than TCP, UDP or an IP protocol number.
The six-octet System ID identifies the router; the NET selector is 00.
A private NET commonly uses AFI 49; plan a unique System ID rather than assuming an IP address is directly encoded as binary.
IS-IS Architecture & NET Addressing
Intermediate System to Intermediate System (IS-IS) is an open-standard, link-state Interior Gateway Protocol (IGP) originally developed by the International Organization for Standardization (ISO) in standard ISO/IEC 10589. Initially designed to route the Connectionless Network Protocol (CLNP) within the Open Systems Interconnection (OSI) protocol suite, IS-IS was subsequently extended in IETF RFC 1195 to route IP alongside CLNP. This variant—known as Integrated IS-IS or Dual IS-IS—is widely deployed across service provider backbones and large-scale enterprise data center fabrics due to its protocol independence, rapid convergence, and clean separation between network transport and routing control.
1. OSI Data Link Encapsulation & Protocol Independence
A defining architectural distinction of IS-IS is its encapsulation layer. While OSPF encapsulates directly within IPv4 packets using IP protocol number 89, and BGP relies on TCP port 179 at the transport layer, IS-IS operates directly at the OSI data link layer (Layer 2).
+---------------------------------------------------------------------------------------------------+
| PROTOCOL ENCAPSULATION STACK COMPARISON |
| |
| OSPFv2 / OSPFv3 Encapsulation: IS-IS Encapsulation: |
| +------------------------------------+ +------------------------------------+ |
| | Layer 2: Ethernet Header | | Layer 2: Ethernet Header | |
| +------------------------------------+ +------------------------------------+ |
| | Layer 3: IP Header (Protocol 89) | | Layer 2 LLC Header (DSAP/SSAP 0xFE)| |
| +------------------------------------+ +------------------------------------+ |
| | OSPF Protocol Data Units (PDUs) | | IS-IS NLPID (0x83) | |
| +------------------------------------+ +------------------------------------+ |
| | IS-IS Protocol Data Units (PDUs) | |
| +------------------------------------+ |
+---------------------------------------------------------------------------------------------------+
On Ethernet networks, IS-IS frames use standard IEEE 802.3 framing with an 802.2 Logical Link Control (LLC) header:
- Destination MAC Address: Dedicated Layer 2 multicast addresses (
01:80:C2:00:00:14for Level 1;01:80:C2:00:00:15for Level 2). - DSAP / SSAP: Set to
0xFE(ISO Network Layer). - Control Field:
0x03(Unnumbered Information). - Network Layer Protocol Identifier (NLPID): Set to
0x83, which uniquely designates IS-IS.
Operational Advantages of Direct Layer 2 Encapsulation
- IP Protocol Independence: Because IS-IS does not require an operational IP address to form neighbor adjacencies or exchange routing updates, it can simultaneously transport IPv4 and IPv6 topologies over common link states using Type-Length-Value (TLV) extensions (such as TLV 128, 130, 232, and 236).
- Security & Resiliency: Because IS-IS does not run over IP, its protocol packets cannot be injected through an ordinary routed IP session. Local attackers and shared CPU/link resource exhaustion remain possible.
- Interface Subnet Independence: IS-IS routers on point-to-point links can form full adjacencies even if their IP interface addresses belong to different subnets or if interfaces are entirely unnumbered.
2. Network Entity Title (NET) & NSAP Addressing Breakdown
Even though Integrated IS-IS routes IP packets, the routing protocol itself requires OSI network addressing for internal node identification. In OSI networking, devices are identified by Network Service Access Point (NSAP) addresses. When an NSAP address identifies the router itself (the network entity) rather than a specific upper-layer service, it is designated as a Network Entity Title (NET).
An NSAP/NET address ranges between 8 and 20 octets (16 to 40 hexadecimal characters) and consists of three distinct functional components:
+---------------------------------------------------------------------------------------------------+
| NETWORK ENTITY TITLE (NET) ADDRESS STRUCTURE |
+-----------------------------------+---------------------------------------+-----------------------+
| AREA ADDRESS (IDP + DSP) | SYSTEM ID | N-SELECTOR |
| (Variable: 1 to 13 octets) | (Fixed: 6 octets / 48 bits) | (Fixed: 1 octet) |
+-----------------+-----------------+---------------------------------------+-----------------------+
| AFI (1 octet) | Area ID (var) | Unique Router Identifier within Area | Must be 00 for Router |
| Example: 49 | Example: 0001 | Example: 1921.6800.0001 (192.168.0.1) | Example: 00 |
+-----------------+-----------------+---------------------------------------+-----------------------+
| Resulting NET: 49.0001.1921.6800.0001.00 |
+---------------------------------------------------------------------------------------------------+
1. Area Address (Initial Domain Part & High-Order DSP)
- Authority and Format Identifier (AFI): The first octet defines the formatting and authority governing the address. The value
49denotes locally administered private addressing (analogous to RFC 1918 private IPv4 space) and is the universal standard in private enterprise deployments. - Area ID: Specifies the logical area to which the router belongs. It can vary from 1 to 12 octets in standard ISO profiles, though enterprise designs typically use a 2-octet area identifier (such as
0001or0002).
2. System ID (Fixed 6 Octets / 48 Bits)
The System ID uniquely identifies the individual router within an area (and across the entire routing domain). To guarantee uniqueness and simplify administration, enterprise network engineers map a router's stable loopback IPv4 address into the 6-octet System ID format by padding each octet to three digits and regrouping into four-digit blocks:
Alternatively, an IEEE 802 MAC address (such as 0010.94ab.0001) can serve as the System ID.
3. N-Selector (SEL, Fixed 1 Octet)
The final octet represents the Network Selector. For host end systems, the SEL identifies an upper-layer protocol or service. For an Intermediate System (router), the N-selector must always be 00. A NET is formally defined as an NSAP address where .
A network engineer configures an IS-IS Network Entity Title (NET) as 49.0001.1921.6800.0001.00 on a Junos router. What is the length of the System ID in this address, and what does the trailing octet represent?
The System ID is 4 octets long, and the trailing octet represents the Level 1 area priority
The System ID is 8 octets long, and the trailing octet represents the OSPF compatibility mode
The System ID is 6 octets (48 bits) long, and the trailing octet is the N-selector which must be 00 for an Intermediate System
The System ID is variable length, and the trailing octet identifies the primary loopback interface logical unit
Sections you finish are checked off in the contents.