The IPv6 header is a fixed 40-byte Layer 3 header designed for faster routing, better scalability and flexible extensions via Extension Headers (RFC 8200).
Facts / key topics
- Standard / RFC: RFC 8200 · OSI layer: Layer 3
- Header length: fixed 40 bytes (no options inside the header)
- Checksum: ❌ none (removed for performance) · Broadcast: ❌ no (replaced by multicast)
- CCNA vs CCNP: CCNA – structure, fields, difference from IPv4; CCNP – extension headers, forwarding, security & troubleshooting
- Where: LAN, WAN, campus, DC, Internet
Why does it exist?
- The IPv4 header had: variable length → slower forwarding · checksum → CPU load · limited address space · options that were ignored or dropped
👉 IPv6 solves this with a simple fixed header + modular extension headers.
How it solves the problem
- Fixed header = fast hardware forwarding
- No checksum → faster processing
- Extension Headers handle fragmentation, security (IPsec) and routing options
- End-to-end design (no NAT requirement)
Step by step
- Packet created: host builds the fixed 40-byte IPv6 header.
- Extension headers are added after the IPv6 header (as a chain).
- Forwarding: router primarily checks Destination Address, Next Header, Hop Limit.
- Errors (CCNP): Hop Limit = 0 → drop · wrong Next Header → drop · MTU problem → ICMPv6 Packet Too Big.
Fields (exam critical)
| Field | Size | Explanation |
|---|---|---|
| Version | 4 bits | Always 6 (4 in IPv4) |
| Traffic Class | 8 bits | QoS (DSCP + ECN) |
| Flow Label | 20 bits | Same flow (QoS / load balancing) |
| Payload Length | 16 bits | Data + extension headers |
| Next Header | 8 bits | Next protocol / extension header |
| Hop Limit | 8 bits | Like TTL in IPv4 |
| Source Address | 128 bits | Sender |
| Destination Address | 128 bits | Receiver |
Key terminology
| Term | Explanation |
|---|---|
| Next Header | Specifies the next header or L4 protocol |
| Flow Label | Identifies a traffic flow |
| Extension Header | Extra functionality without changing the base header |
| Hop Limit | Maximum number of hops |
Extension Headers (CCNP focus)
Typical order:
- Hop-by-Hop Options
- Destination Options
- Routing
- Fragment
- Authentication (AH)
- ESP
- Upper layer (TCP/UDP/ICMPv6)
⚠️ Fragmentation happens only on the source host – never on routers.
Dependencies & special considerations
- IPv6 MTU: minimum 1280 bytes
- Fragmentation: host only
- Security: IPsec is architecturally “built-in”
- Performance: some devices drop packets with many extension headers
Design & best practices
- Avoid unnecessary extension headers · monitor ICMPv6 (don’t block blindly) · use Flow Label consistently in DC / QoS designs
Exam tips ⚠️
- “IPv6 header length?” → ALWAYS 40 bytes
- “TTL in IPv6?” → it’s called Hop Limit
- “Options in the IPv6 header?” → ❌ no (Extension Headers!)
- “Fragmentation on a router?” → ❌ never
30-second summary
- Fixed 40 bytes · no checksum, no broadcast · Next Header replaces “Protocol + Options” · extension headers = flexibility · designed for fast forwarding
| Level | Expectation |
|---|---|
| CCNA | Structure, fields, IPv4 vs IPv6 |
| CCNP | Extension headers, MTU, security, troubleshooting |