Transcript Overview of Scalable Networks

1 of 35
Single Area OSPF Concepts
2 of 35
OSPF Basics
3 of 35
OSPF versus RIP
• Open Shortest Path First
(OSPF), like RIP, is based on
“Open” standards.
 RFC 2328
• OSPF is often preferred over
RIP because of its scalability.
 It can be configured on
smaller networks using one
“Area” (shown in yellow)
 Or scaled to larger networks
with virtual no limit.
 Networks and Areas are
easily added or
removed
 VLSM and Route Summaries
are fully supported.
4 of 35
OSPF versus RIP
• Open Shortest Path First
(OSPF), like RIP, is based on
“Open” standards.
 RFC 2328
• OSPF is often preferred over
RIP because of its scalability.
 It can be configured on
smaller networks using one
“Area” (shown in yellow).
 Or scaled to larger networks
with virtually no limit.
 Networks and Areas are
easily added or
removed,
 VLSM and Route Summaries
are fully supported.
5 of 35
OSPF versus RIP
• Open Shortest Path First
(OSPF), like RIP, is based on
“Open” standards.
 RFC 2328
• OSPF is often preferred over
RIP because of its scalability.
 It can be configured on
smaller networks using one
“Area” (shown in yellow).
 Or scaled to larger networks
with virtually no limit.
 Networks and Areas are
easily added or
removed.
 VLSM and Route Summaries
are fully supported.
6 of 35
OSPF versus RIP
• Open Shortest Path First
(OSPF), like RIP, is based on
“Open” standards.
 RFC 2328
172.16.4.0/22
• OSPF is often preferred over
RIP because of its scalability.
172.16.0.0/22
172.16.0.0/19
 It can be configured on
smaller networks using one
“Area” (shown in yellow).
 Or scaled to larger networks
with virtually no limit.
 Networks and Areas are
easily added or
removed.
172.16.8.0/22
172.16.0.0/16
172.16.64.0/19
172.16.64.0/22
172.16.72.0/22
172.16.68.0/22
172.16.32.0/19
172.16.32.0/22
172.16.40.0/22
172.16.36.0/22
 VLSM and Route Summaries
are fully supported.
7 of 35
OSPF versus RIP
• Speed of Convergence
 RIP broadcasts every 30 seconds
8 of 35
OSPF versus RIP
• Speed of Convergence
 OSPF is event-driven.
 Only changes are flooded to other routers.
 Sends an LSA when a change occurs.
9 of 35
OSPF versus RIP
• Path Selection
 RIP can pick suboptimal paths because only hops are
considered.
 OSPF calculates the “Cost” of each link, which defaults to
bandwidth (108/bps).
10 of 35
Administrative Distance
• Because OSPF has
superior convergence
and path selection, it
has a lower AD than
RIP.
 Remember: AD is a
measure of a route
source’s
“trustworthiness” or
“believability”
Route Source
Default
Administrative
Distance
Connected interface
0
Static route out an interface
0
Static route to a next hop
1
EIGRP summary route
5
External BGP
20
Internal EIGRP
90
IGRP
100
OSPF
110
IS-IS
115
RIP (v1 and v2)
120
EGP
140
External EIGRP
170
Internal BGP
200
Unkown
255
11 of 35
OSPF Terminology
 Links—networks a router knows about; each router interface is a
“link”.
 Area—a group of routers identified with a unique ID; all routers in the
same area share the same link-state database.
 Cost—is the bandwidth of the media; can be manually configured.
 SPF Algorithm (Dijkstra)—calculated by each router to choose the
lowest-cost path.
 Link-state—is a link “up” or “down”?
 LSA—a link state advertisement.
 Adjacencies Database—keeps track of all directly connected routers
(also called neighbors).
 Link-State Database—also known as the Topology database; picture of
who is connected to what; all routers should have the same L-S DB.
 Forwarding Database—known as the Routing table where the lowestcost paths are installed.
 Designated Router/Backup Designated Router (DR/BDR)— routers that
are elected on multiaccess networks to be the focal point for routing
updates.
12 of 35
OSPF Terminology Graphic
13 of 35
OSPF Packet Types
14 of 35
OSPF Packet Types
• OSPF uses a variety of packets to communicate
with neighbors and the DR/BDR.
 Type 1: Hello; a 64-byte packet sent at regular intervals
to keep a link “alive”.
 Type 2: DBD (Database Description); summary
contents of a router’s link-state database sent to a
newly discovered neighbor.
 Type 3: LSR (Link-State Request); requests more
specific information about a link from a neighbor’s
link-state database.
 Type 4: LSU (Link-State Update); transports LSAs to
neighbor routers; for example, a reply to an LSR.
 Type 5: LSAck (Link-State Acknowledgement);
acknowledges receipt of an LSA; OSPF’s routing
updates are connection-oriented.
15 of 35
OSPF Packet Header (All Types)
• Below is the 20 byte packet header appended to the front of all
OSPF packets.
 Version specifies OSPF version; routers must be running the same
version or adjacency with neighbors cannot be established.
 Type specifies packet type (Type 1, Type2, etc.)
 Packet Length is the length of the entire OSPF packet in bytes,
including the standard OSPF packet header.
 Router ID is the IP identity of the router who is originating the
packet.
 Area ID is the OSPF area that the packet is being sent into.
 Authentication, if configured, is specified.
0
8
Version
16
Type
24
31
Packet Length
Router ID
Area ID
Checksum
Authentication Type
Authentication Data
16 of 35
Hello Packet Header (Type 1)
• Below is the additional fields added to the OSPF packet
header to make an OSPF Hello packet header.
 Network Mask is the number of bits turned on in the Subnet
mask used by the sending router’s Router ID.
 Hello Interval is number of seconds between sending
router’s hellos
 10 sec. or 30 sec., depending on network type
0
8
16
24
31
Network Mask
Hello Interval
Options
Router Priority
Dead Interval
Designated Router
Backup Designated Router
Neighbor Router ID
(additional Neighbor Router ID fields, if necessary)
17 of 35
Hello Packet Header (Type 1)
 Options are described in Section A.2 of RFC 2328
 Router Priority is used for DR/BDR Elections. If set to 0,
sending router is ineligible to become Designated Router.
 Dead Interval is the number of seconds before sending
router will consider a silent neighbor to be down.
 Defaults to 4 times the Hello Interval
0
8
16
24
31
Network Mask
Hello Interval
Options
Router Priority
Dead Interval
Designated Router
Backup Designated Router
Neighbor Router ID
(additional Neighbor Router ID fields, if necessary)
18 of 35
Hello Packet Header (Type 1)
 Designated Router is the Router ID of the DR for this
network, in the view of the sending router.
 Backup Designated Router is the Router ID of the BDR for
this network, in the view of the sending router.
 Neighbor Router IDs are the Router IDs of each router from
whom valid Hello packets have been seen recently within
the Dead Interval.
0
8
16
24
31
Network Mask
Hello Interval
Options
Router Priority
Dead Interval
Designated Router
Backup Designated Router
Neighbor Router ID
(additional Neighbor Router ID fields, if necessary)
19 of 35
OSPF Operation
20 of 35
OSPF States
• OSPF neighbor
adjacencies are
established through a
seven step process:
 Down
 Init
 2Way
 ExStart
 Exchange
 Loading
 Full
• Each state is discussed
in following slides.
21 of 35
Down, Init, 2Way States
• When a router first comes online, it is in the Down State and
begins sending Type 1 Hello packets.
• When another router hears the new router’s Type 1 Hello
packet for the first time, it will enter the Init State.
• Once the new router sees its own ID in the Hello packet sent
by the neighbor, the routers move to the 2Way State.
22 of 35
ExStart State
• The routers now enter the ExStart State.
• Router priority or Router ID is used to
determine master/slave relationship
23 of 35
Exchange State
• During the Exchange State, Type 2 DBD packets
are exchanged.
 These are a summary of each router’s Link-State
Database.
 After receiving a Type 2 DBD, the router sends an
acknowledgment.
24 of 35
Loading State
• The Loading State is used only if one or more
routers in the unconverged network “heard new
information.”
 The router will request more info; Type 3 LSR
 The receiving router sends an update; Type 4 LSU
25 of 35
Full State
• Routers enter Full State and can now both
calculate the SPF algorithm (Dijkstra) in
parallel.
26 of 35
Steps in the OSPF Operation
• Routers move through
five distinct steps of
operation.
 Step 1: Establish Router
Adjacencies
 Step 2: Elect a DR and
BDR
 Step 3: Discover Routes
 Step 4: Select
Appropriate Routes
 Step 5: Maintain
Routing Information
• Each Step is discussed
on following slides.
27 of 35
Step 1: Discover Neighbors & Establish Adjacencies
• This process—
already discussed
previously in the
description of Init
and 2Way states—is
similar for
broadcast
networks.
28 of 35
Step 2: Elect a DR and a BDR
• Occurs during
the ExStart State:
 The router with
the highest
priority or
highest
configured IP
address
becomes the DR.
 The BDR is
chosen the same
way.
29 of 35
Step 3: Discover Routes
• Type2 DBDs are
now exchanged
 Although the
graphic does not
show it, the BDR
silently listens to
all traffic.
30 of 35
Step 4: Select Appropriate Routes
• The SPF
Algorithm is now
calculated in
parallel with
every other
router in the
Area.
31 of 35
Step 5: Maintain Routing Information
• Exchange periodic
Hellos to detect
changes in status of
neighbors.
32 of 35
Network Types
Network Type
Characteristics
DR/BDR Elected?
Ethernet, Token Ring, FDDI
Yes
Frame Relay, X.25, SMDS
Yes
PPP, HDLC
No
Configured by administrator
with subinterfaces
No
Broadcast Multiaccess
Nonbroadcast Multiaccess
Point-to-Point
Point-to-Multipoint
33 of 35
Down
Init State
Multicast Hello
packets
Listening routers add
the new router to the
adjacency table
Routers reply to Hello
packets with their
own Hello packet
2Way State
Choose
DR/BDR
Link type is
multiaccess
Adjacencies must be
established (depends on
link type)
Link type is
point-to-point;
determine
“master/slave”
Compare all
Router Priority
values
Is there a tie?
Originating router adds
all replying routers to
adjacency table
ExStart State
Yes
Compare
Router IDs
No
Take highest
value
Assign as DR
Take secondhighest value
Assign as BDR
Exchange
link-state
information
Exchange
Any final LSAs
are also
exchanged
Loading
Exchange Hello packets
at intervals to maintain
updated routing
information
Full State
34 of 35