Unicast Routing in IPv6

Babu Ram Dawadi

Unicast Routing Protocol
 Unicast routing is a process that enable sender to send an unicast IP
packets to the destination node.

 1 router or more intermediate routers may be used, depending to the

destination of the node. (Figure 1)

 Unicast routing protocol is a set of rules of forwarding unicast traffic from

a source to a destination on an internetwork.

S D S D The router is using only

1 port to forwards the
` received unicast packet
S

Source (S)
D

`
S D S D

Destination (D)

Fig. 1. Unicast Routing

Unicast Routing Protocol
 Unicast Routing Protocol consists of:
– RIP (Routing Information Protocol)
– OSPF (
– BGP

 They each serve a different purpose.

Routing

Interior Exterior

RIP OSPF BGP

Fig. 2. Types of Unicast Routing Protocol
RIPng (RFC 2080)
 Based on Distance Vector algorithm known as
bellman ford algorithm

 Router keeps the following entries in the routing table

– IPv6 Route
• Address prefix and prefix length of the destination address
– Next Hop Address
• The IPv6 Address (link-local) of the first router along the path
– Next Hop Interface
• The physical interface used to reach the next hop
– Metric
• Number indicating the total distance to the destination. RIPng
advertizes directly connected routes with the configured outgoing
metric of 1
RIPng routing table
 Timer
– Amount of time since the information about the route
was last updated

 Route change flag

– Set to control triggered routing updates

 Route Source
– Entity to provide route information eg: Ripng, OSPF etc..
RIPng Message Format
 UDP based protocol using udp port number 521

IPv6 UDP RIPng RTE1 RTE2 RTE3 ……… RTEn

Header Header Header

20 Bytes/RTE

1: Request  Command 1: ask system to send all

Command 1B 2: Response or part of its routing table

Version 1B  Command 2: sends an update

message containing all or parts of
the senders routing table.
Unused 2B
Routing Table Entry (RTE)
 RIPng header is followed by one or more routing
table entries (format of Routing table entry)
16 B …. IPv6 Prefix

2B Route Tag

1B Prefix Length

1B Metric(1-16)
RTE..
 Route Tag
– It may be used to carry additional information about a route
learned from another routing protocol eg: BGP
– The number of RTEs within single updates depends on the
MTU of the medium between two neighboring routers
– No of RTEs=[INT(MTU-IPv6 Hdr len-UDP Hdr len -RIPng Hdr len) / RTE-Size]

 Timers
– RIPng uses different timers to control updates of the routing
information
– Update timer
• By default, every 30 seconds, RIPng process wakesup on each
interface to send an unsolicited routing response to the neighboring
routers
Timer
 Timeout Timer
– Each time a route entry is updated and the timeout timer
is reset to zero
– If the route entry reaches 180 secs (default), without
another update, it is considered to have expired, metric
set to 16 and garbage collection process starts

 Garbage collection timer (hold down timer)

– Set to 120 secs (default) that have timeout or been
received with a metric of 16 after expiration, the route
entry finally be removed from the routing table.
OSPF
 The difference between intra-domain and inter-domain (Fig. 2.)
– Routing within the same AS (Autonomous System) is referred to as intra-
domain.
– Routing in the different AS (Autonomous System) is referred to as inter-
domain.

 OSPF is an interior routing protocol

– Intra-domain routing protocol

AS 1 Router 1 AS 2

Router 2

a) Intra-domain b) Inter-domain
Fig. 3. Interior Routing Protocol
OSPF Header
IPv6 Header OSPF Header OSPF Message
NH: 89
40Byte 16 Byte

Version 1B RouterID 4B

AreaID 4B
Packet
Type 1B Checksum 2B
Packet
Length 2B Instance ID 1B

Unused 1B
OSPF Process
 Link State routing
– Each node within the autonomous system has the
information about the entire topology.
– Each node in the domain build up the routing table
using Dijkstra’s algorithm.

 Link State Database (LSDB) contains link state advertisement is

send to every router in the same domain.
– Each router will be updated with the latest copy of LSDB

 Based on the LSDB, router creates a Shortest Path First (SPF) tree
– Using Dijkstra’s a Algorithm

 A routing table can be derived from the SPF tree which contains
the best route to each router.
Types of OSPF Packets
OSPF Packets

Type 1 Type 2 Type 3 Type 4 Type 5

Database Link State Link State Link State

HELLO Acknowledgement
Description Request Update

Inter-Area-Prefix Inter-Area-Router Group-

Router Link Network Link AS-External-LSA
LSA LSA Membership-LSA
0x2001 0x2002 0x4005
0x2003 0x2004 0x2006

Type-7-LSA Link-LSA Intra-Area-Prefix

0x2007 0x0008 LSA
0x2003

Fig. 4. Types of OSPF Packet

How do OSPF routers on the same
link establish adjacencies?

What does it mean to be

adjacent? GWE.
OSPF- Forming Adjacencies
 OSPF routers on the same link establish adjacencies
– Using Hello packet

 An OSPF router need to go through 7 steps from no connection to

full adjacency when it is first initialized.
– Down State
– Init State
– Two Way state
– Exstart
– Exchange State
– Loading State
– Full Adjacency
OSPF- Forming Adjacencies (cont)
Down State No information has yet None
been exchanged.

Init State Routers send hello packets at regular Hello Packet

intervals to establish relationships. (Type 1)

Two-way A router sees itself in a hello packet. Hello Packet

State (Type 1)

Routers negotiate master/slave relationship by Hello Packet

Exstart (Type 1)
comparing their router id using hello packets.

Exchange Database Description

Neighbors start communicating their link-state
State (Type 2)
information with the others.

Loading The router has the initial information of each route, Link State Request
State they may request for more complete information. Link State Update
Link State Acknowledgement
(Type 3,4,5)

Full All
Routers are fully adjacent. (Type 1,2,3,4,5)
OSPF- Forming Adjacencies (cont)
Hello Packets from router neighbour X

 Processing of
Authentication
AreaID
(Accepted)

Hello packet
YES

Am I already Reset timer of

NO
Hello Interval Discard adjacent to X? Neighbour X
NO
match? packet

YES
YES NO

Neighbour X
RouterDead Discard Neighbour X sees
NO NO changes to the
Interval Match? packet me as neighbour?
INIT State

YES

Neighbour X
E- and N- Discard changes to the 2-
NO
bit match packet way state

YES
YES

Incoming Transit
interface of type?

Is nieghbour X or Neighbour X
NO
am I DR/BDR stays 2 ways

Point to Point
Fig. 5. Processing a Hello Packet
YES
(Reference: IPv6 Essential Pg.169)
Forming Adjacency with Router X
OSPF- Forming Adjacencies (cont) The first OSPF neighbor is
Down state. It means that no
information (hellos) has been
received from this neighbor.
But it still can send Hello Router 2 has received a
message to its neighbours hello packet from its
Router1 has seen the Router neighbor, butreceived
the
Point to Point (No DR,BDR) Router2 also has
2’s hello packet which had
 Forming an adjacency Transit Link (DR,BDR, or DROther)
the Router 1’s helloID
receiving router's was
packet
included its own Router ID not included in the hello
which had also included its
within the received hello
– Phases packet's neighbor field
Router 1(R1) packet
Router 2 (R2)own Router ID within the
received hello packet's
– Packet type involved Hello (Neighbour = 0)
Init neighbor field
Hello (Neighbour = R1)
2-Way
Hello (Neighbour = R2)
At this state, a router decides whether 2-Way
DD (Seq = x, Init, More, Master)
to become adjacent with this neighbor. Ex-Start Ex-Start
DD (Seq = y, Init, More, Master)
The decision on adjacencies is always DD (Seq = y, More, Slave)
depend on the link media. (Types of Exchange Exchange
network) DD (Seq = y+1, More, Master)

DD (Seq = y+1, More, Slave)

Point to Point
-No DR/BDR is needed or used because
Time

because only 2 neighbour routers. DD (Seq = y+n-1, More, Master) no more to send

DD (Seq = y+n-1, More, Slave)

Broadcast Multiple Access
DD (Seq = y+n, More, Master)
-DR and BDR are elected.
DD (Seq = y+n, More, Slave) no more to send
Non-Broadcast Multiple Access Link State Request
-DR and BDR are elected Loading Loading
Link State Update

Link State Acknowledgement

Link State Request

Link State Update

Fig. 6. Forming an Adjacencies – Part I Full

Link State Acknowledgement
Full
Hello
(Reference: IPv6 Essential Pg.170)
OSPF- Forming Adjacencies (cont)
 Types of Network:

Broadcast Multiple Access

Ethernet, Token Ring

Frame Relay

Point to Point
PPP, HDLC Non Broadcast Multiple Access (NBMA)
X.25, Frame Relay
Fig. 7. Types of Network
OSPF- Forming Adjacencies (cont)
 When DR and BDR are elected, we can say that the adjacencies
was established.

 The DR and BDR are elected based on several criteria

– DR – with highest router ID
– BDR – with second highest router ID
– Priority 0 will never be DR or BDR
– If priorities are same, the higher Router ID is elected.
– If DR fails, BDR becomes DR, and new BDR is elected.

 After adjacent, the routers already created reliable channels to

their neighbors.

 These reliable channels is important for the routers to exchange

Link State Database (LSDB) with the neighbors..
How do OSPF routers synchronize
their Link State Databases?
OSPF - Link State Databases
Point to Point (No DR,BDR)
Transit Link (DR,BDR, or DROther)
The Router 1 is
 When the routers have elected DR attempted to start the
Router 1(R1)
But Router 2 will
Router 2 (R2)
Exchange because reply “No”. I will
and BDR, the databases need to “he” might thought Hello (Neighbour = 0) start the Exchange
that “he” has the Init
get synchronized Hello (Neighbour = R1) State first, because
highest Router ID 2-Way I have a higher
– Routers enter ExStart After the election of Master Hello (Neighbour = R2)
2-WayRouter ID
and Slave, the routers start
– Decide who is the Master and Slave.database descriptor
exchange Ex-Start
DD (Seq = x, Init, More, Master)
Ex-Start
(DBD) packets. DD (Seq = y, Init, More, Master)
– Router with the Highest ID will
!But your DR/BDR
become Master and only this router Exchange DD (Seq = y, More, Slave)
Exchange

! can increment the sequence number

It– might logic toelection
election might be purely
The Master/Slave conclude Each DBD packet has a
is onsequence
a number
DD (Seq = y+1, More, Master)

DD (Seq = y+1, More, Slave)

per-neighbor basis which can be DBD contain link-

that the DR/BDR withInthe
based on the incremented
the only by
DD (Seq = y+n-1, More, Master) no more to send state advertisement
– Master/Slave election is irrelevant
this state,
toand (LSA) headers only
actual exchangeexplicitly
master
highestDR/BDR
Router ID for
election will become
a network.
of DD (Seq = y+n-1, More, Slave)
and describe the
higher priority occurs of
the master during this process
linkacknowledged by slave
state information DD (Seq = y+n, More, Master)
contents of the
DD (Seq = y+n, More, Slave) no more to send entire link-state
configure
master/slave election. Loading
Link State Request database
Loading
Routers are fully
on router adjacent
Link State Update
Based
with eachon the
Link State Acknowledgement
information
other. All the router and The neighbor
network LSAsprovided
are by the then provides the
exchanged and therouters
DBDs, requested link-
Link State Request
send link-state
routers' databases are state information
request packets
fully synchronized. Link State Update in link-state
Link State Acknowledgement update packets
Full Full
Fig. 7. Forming an Adjacencies-Part II Hello

(Reference: IPv6 Essential Pg.170)

OSPF – Database Description (DBD) Packet

OSPF Packet Type 2

Database Description Packet

Database Description fields:

I bit = Init bit
M bit = More bit
MS bit = Master / Slave bit
OSPF – Database Description (DBD) Packet

I bit is set to1

Indicate that this is First DBD packet send by this router 172.16.51.3
This packet contain no data.

M bit is set to 1
Indicate that there are more DBD packets to follow

MS bit is set to 1
Indicate that this router (172.16.51.3) declares itself to be Master
OSPF – Database Description (DBD) Packet

I bit is set to 1
Indicate that this is First DBD packet send by this router 202.249.25.234
This packet contain no data.

M bit is set to 1
Indicate that there are more DBD packets to follow

MS bit is set to 1
Indicate that this router (202.249.25.234) declares itself to be Master

Empty
OSPF – Database Description (DBD) Packet

I bit is set to 0
Indicate that this is NOT First DBD packet send by this router 172.16.51.3.
It contain database description of router 172.16.51.3

M bit is set to 0
Indicate that there are no more DBD packets to send
All database descriptions have been sent

MS bit is set to 0
Indicate that this router (172.16.51.3) declares itself to be Slave
OSPF – Database Description (DBD) Packet

I bit is set to 0
Indicate that this is NOT First DBD packet send by this router 202.249.25.234
It contain database description of router 202.249.25.234

M bit is set to 1
Indicate that there are more DBD packets to send
The database descriptions will still be sent

MS bit is set to 1
Indicate that this router (202.249.25.234) is still a Master
OSPF – Database Description (DBD) Packet

The router 172.16.51.3 is still sending

a empty packet even it had no data
description to send. This empty packet
I bit is set to 0
is help to keep thethatDBD
Indicate sequence
this is NOT First DBD packet number
send by this router 172.16.51.3.
It contain database description of router 172.16.51.3
matched because router 202.249.25.234
is still sending thethatDBD
Indicate there arepacket.
M bit is set to 0
no more DBD packets to send
All database descriptions have been sent

MS bit is set to 0
Indicate that this router (172.16.51.3) is still a Slave

Empty
OSPF – Database Description (DBD) Packet

I bit is set to 0
Indicate that this is NOT First DBD packet send by this router 202.249.25.234
It contain database description of router 202.249.25.234

M bit is set to 0
Indicate that there are no more DBD packets to send
All database descriptions have been sent

MS bit is set to 1
Indicate that this router (202.249.25.234) is still a Master
OSPF – Database Description (DBD) Packet

I bit is set 0
Indicate that this is NOT First DBD packet send by this router 172.16.51.3.

M bit is set 0
Indicate that there are no more DBD packets to send
All database descriptions have been sent

MS bit is set 0
Indicate that this router (172.16.51.3) is still a Slave

Empty
OSPF – Link State Request Packet

OSPF Packet Type 2

Link State Request Packet

Multiple requests can be sent

using a single packet.
(Reference IPv6 Essential pg.172)
OSPF – Link State Update Packet (0x2001)

OSPF Packet Type 4

Link State Request Packet

Link State Update Packet Type 0x2001

Router-LSA
OSPF – Link State Update Packet (0x2002)

OSPF Packet Type 4

Link State Request Packet

Link State Update Packet Type 0x2002

Network-LSA
OSPF – Link State Update Packet (0x4005)

OSPF Packet Type 4

Link State Request Packet

Link State Update Packet Type 0x4005

AS-External-LSA
OSPF – Link State Update Packet (0x0008)

OSPF Packet Type 4

Link State Request Packet

Link State Update Packet Type 0x0008

Link-LSA

Purpose: List all IPv6 Prefix attached to the link

OSPF – Link State Update Packet (0x2009)

OSPF Packet Type 4

Link State Request Packet

Link State Update Packet Type 0x2009

Intra-Prefix-LSA
OSPF – Link State Acknowledgement Packet

OSPF Packet Type 5

Link State Acknowledgement Packet
Learn how to see the above processes
using ospfd, and ospf6d.
Example: OSPF Interface Information

This command shows the

interface configuration
parameter such as costs,
priority, DR/BDR for this
interface, and Status.
Example: Neighbours

This command shows all

neighbors connected to
the node and their status.
Example: Link State Database

The type of LSA can be specify

from the command and these is
the 8 types of the LSA packet.
Example: Link State Database (cont)

This is the LSA database

summary.
Example: Link State Information
Example: Shortest Path Tree

This is the SPF Tree

information. The SPF tree
is used to calculate the
shortest path from each
node to all other nodes in
the area.
OSPF6D - Troubleshooting
Case 1 : Unable telnet to OSPF

Where is ospf6d process?

The telnet to OSPF failed because the process

of OSPF6D was not running
Case 1 : Successfully Telnet to OSPF

ospf6d process

The telnet to OSPF success because

the process of OSPF6D was running
Case 2: Unable to enter Full State with DR in OSPF

Changing the rl0 MTU size to

be 1500 bytes for OSPF6D
Case 2: Unable to enter Full State with DR in OSPF

Adjacency always stay at ExStart State

and will never proceed to Full State
Case 2: Unable to enter Full State with DR in OSPF

From the ospf6d.log file,

we can see that the
adjacency with DR was
stuck at Exstart State
Case 2: Successfully to enter Full State with DR

Changing the rl0 MTU size to

be 1452 bytes for OSPF6D
Case 2: Successfully to enter Full State with DR

From the ospf6d.log

we can see that now
the adjacency between
this router and DR can
proceed till Full State
Case 2: Successfully to enter Full State with DR

Now it can proceed till Full State

Case 3 : Router 172.16.51.3 DOWN

The 172.16.51.3 doesn’t send

Hello packet, the router Dead
Timer was counting down.

Router Dead Timer for

172.16.51.3 was times up.
This router was wiped out
from neighbors list
Case 4 : Router 172.16.51.3 UP

The 172.16.51.3 send

Hello packet, and received by
this router. Adjacency between
this 2 router running again.

Due to the parameter configuration

of this 2 router, both don’t want
to be DR/BDR, thus they stay at
Twoway/DRother state
Thank you