3Com Router User Manual

							Configuring OSPF427
hello packet before this interface sets up neighboring relations with the adjacent 
routers. 
The interface can be configured into nbma mode on the broadcast network 
without multi-access capability.
If not all routers are inter-reachable on NBMA network, the interface can be 
configured into 
p2mp mode.
If the router has only one opposite terminal in NBMA network, the interface can 
also be changed to 
p2p mode.
The difference between an NBMA network and a point-to-multipoint network 
includes the following distinctions:
■In the OSPF protocol, NBMA refers to those as fully connected, nonbroadcast 
and multi-access networks. But point-to-multipoint network does not 
necessarily require full connection.
■DR and BDR should be elected on NBMA while there is no DR or BDR on 
point-to-point network.
■NBMA is a default network type. For example, if the link layer protocol is X.25 
or frame relay, OSPF regards the network type of this interface as NBMA 
(whether the network is wholly connected). Point-to-multipoint is not a default 
network type. No link layer protocol can be considered as a point-to-multipoint 
protocol because it must be a modification from other network types. The most 
common practice is to change the not fully connected NBMA to a 
point-to-multipoint network.
■An NBMA network sends messages in unicast mode and the peer must be 
configured manually. In point-to-multipoint network, messages are sent either 
in unicast mode or in multicast mode.
Perform the following configurations in interface view.
Ta b l e 490   Configure the Network Type of the OSPF Interface
After a new OSPF network type is configured, the old network type on the 
interface will be replaced automatically.
Configuring Sending 
Packet CostYou can configure the cost of sending a packet on the interface, otherwise OSPF 
automatically calculates the cost value according to the baud rate of the current 
interface.
Perform the following configurations in interface view.
Ta b l e 491   Configure Sending Packet Cost
OperationCommand
Configure the network type of the OSPF 
interfaceospf network-type { broadcast | nbma 
| p2mp | p2p
 }
Delete the specified OSPF network typeundo ospf network-type { broadcast | 
nbma | p2mp | p2p
 }
OperationCommand 
						

							428CHAPTER 29: CONFIGURING OSPF
The default value of the cost of sending a packet on the interface is calculated 
automatically according to the interface baud rate as follows:
The default value is automatically calculated according to interface baud rate.
■If the baud rate is less than 2000 bps, 2000 is taken, and the overhead value is 
100000000/2000=50000. 
■If the baud rate is greater than 100000000 bps, 100000000 is taken, and the 
overhead value is 100000000/100000000=1. 
■If the baud rate is between 2000 bps and 100000000bps, the overhead value 
is the result of 100000000/interface rate.
Configuring a Peer for 
the NBMA InterfaceSpecial configuration is needed for the network of an NBMP interface. Since the 
adjacent router cannot be found by broadcasting hello packets, the IP address of 
the adjacent router should be specified manually for the interface, as well as 
whether the adjacent router has a voting right. This is specified with the 
ospf 
peer ip-address [eligible]
 command. To use the eligible attribute, this 
adjacent router must have no voting right.
On X.25 and frame relay networks, you can configure a map to make the whole 
network fully connected so there is a virtual circuit between any two routers ont 
eh network and they are directly reachable. Then OSPF can process like a 
broadcast network. The IP address of the adjacent router, and whether it has a 
voting right, must be specified manually for the interface because the adjacent 
router cannot be found dynamically by broadcasting hello packets.
Perform the following configurations in interface view.
By default, no peer of the NBMA interface is specified.
When you configure the peer of the NBMA interface, the following items are 
necessary:
■The configured hello timer and dead timer between neighbors must be 
identical.
■The configured link route type betwen neighbors must be identical.
■The area number that neighbors belong to must be identical.
■The authentication mode (simple text or MD5), authentication password, and 
key-id of the area that neighbors belong to must be identical.
■The configured stub attribute of the areas including the neighbors must be 
consistent.
Configure sending packet costospf cost cost
Reset the sending packet costundo ospf cost
Ta b l e 492   Configuring a Peer for the NBMA Interface
OperationCommand
Configure a peer for NBMA interfaceospf peer ip-address [eligible]
Cancel or delete a peer for NBMA interfaceundo ospf peer ip-address 
						

							Configuring OSPF429
Specifying the Router 
PriorityIt is necessary to establish the peer relationship manually between interfaces for 
multi-point access network, (NBMA and broadcast type networks). But 
establishing peer relationshipoccupies large amounts of system resources when 
there are hundreds of routers in the network. To resolve this issue, OSPF specifies a 
“designated router” (DR). All routers within the same network segment send the 
relationship information to the DR, which broadcasts the link status of each 
network segment. In this way, the number of the peer relationships between 
different routers on the multi-access network is significantly reduced.
The priority of a router interface determines the qualifications of the interface in 
voting for the DR. The interface with a higher priority is considered first when the 
voting rights conflict.
The DR is not designated manually, but voted by all routers in the local network 
segment. The routers of Priority>0 in the local network segment can be used as 
the “candidates”. The router with the greatest priority value is selected among all 
routers that claim to be DR. If two routers have the same priority, the one with 
greater router ID is selected. Routers vote by Hello packet. Each router writes the 
DR into the Hello packet and sends it to all other routers on the network segment. 
When two routers in the same network segment claim to be the DR, the one with 
the higher priority is chosen. If the priorities are equivalent, the one with higher 
router ID is chosen. If the priority of a router is 0, it is not selected as the DR or 
“backup designated router” (BDR).
If a DR fails due to a specific fault, a new DR must be elected, with 
synchronization. This can take a long time, during which, the route calculation is 
not correct. To shorten the process, OSPF puts forward the concept of the 
“backup designated router” (BDR). The BDR is actually a standby for the DR and is 
voted together with DR. The BDR also creates relations with all neighboring 
routers in the network segment and exchanges routing information with them. 
When the DR fails, the BDR becomes the DR immediately without the need for 
re-election. Because the neighboring relationship is already created, this takeover 
process is instantaneous. Of course a new BDR needs to be elected again but 
during the election, the route calculation is not affected.
It should be noted that:
■The DR in the network segment is not necessarily the router with the highest 
priority. Similarly, the BDR is not necessarily the router with the second highest 
priority.
■The DR is a role in a single network segment, based on the router interface. A 
router can be a DR on one interface and a BDR or DROther on another 
interface.
■The DR is elected on a broadcast interface or NBMA interface. It is not 
necessary on a point-to-point interface or point-to-multipoint interface.
Perform the following configurations in interface view.
Ta b l e 493   Specify the Router Priority
OperationCommand
Set the priority of the interface when 
selecting a designated routerospf dr-priority value
Return to the default router priority undo ospf dr-priority 
						

							430CHAPTER 29: CONFIGURING OSPF
Specifying the Hello 
IntervalThe Hello packet is periodically sent to the neighboring router to find and maintain 
OSPF neighbor relationship, and to elect the DR and BDR in the NBMA and 
broadcast networks. When one router is started, it only sends hello packets to the 
neighbors whose precedences are larger than 0, that is, the routers can possibly be 
elected as DR or BDR. You can configure the interval for sending hello packets. If 
the interval is too short, the network change can be easily found but the network 
load will be greatly increased. An appropriate value must be selected for the 
specific network conditions. 
On NBMA and point-to-multipoint networks, the poll-interval attribute should be 
configured to specify the period of sending polling hello packet before this 
interface sets up a neighbor relationdship with the adjacent routers.
The poll-interval attribute must be at least 3 times the value for hello-timer.
Perform the following configurations in interface view.
Ta b l e 494   Specify Hello Intervall
By default, the hello-timer on the p2p interface is 10 seconds and the hello timer 
on the 
p2mp amd nbma interfaces on the same network segment must be 
identical.
Specifying the Dead 
IntervalThe expiration time of a neighboring router means that if a hello packet of the 
neighbor router (peer) is not received within a certain period, the neighbor router 
is invalid. You can specify the dead-timer, the period where the peer route fails. 
The value of the dead-timer must be at least 4 times the value of the hello-timer.
Perform the following configurations in interface view.
Ta b l e 495   Specify Dead Interval
By default, the dead-timer on the p2p interface is 40 seconds and on p2mp and 
nbma
 interface the dead-timer is 120 seconds, ranging from 1 to 65535 seconds.
Note that: 
■The dead-timer of the router on the same network segment must be the same.
■When you modify the network type, the hello-timer and dead-timer are both 
restored to their default values.
OperationCommand
Set the time interval for the interface to 
send hello packetsospf timer hello seconds
Return to the default hello interval timeundo ospf timer hello
Specify the length of poll-interval on 
NBMA and point-to-multipoint network 
typeospf timer poll seconds
Return to the default poll interval timeundo ospf timer poll 
OperationCommand
Specify the expiration duration of the 
OSPF neighborospf timer dead seconds
Return to the default value of dead 
intervalundo ospf timer dead 
						

							Configuring OSPF431
Specifying the 
Retransmitting IntervalThe router waits for confirmation from the neighbor to whom it has sent an LSA. If 
the router does not receive the neighbors confirmation after a specified interval, 
the retransmitting interval, it resends the LSA. You can set the time interval for 
re-transmitting an LSA.
Perform the following configurations in interface view.
Ta b l e 496   Specify Retransmitting Interval
By default, the retransmitting interval is 5 seconds.
The retransmitting interval must be twice of the period when a message is 
transmitted between two routers.
The interval for retransmitting an LSA between adjacent routers must not be so 
small as to cause unnecessary retransmission.
Specifying the 
Transmit-delayThe LSA ages in the link status database (LSDB) of the local router (1 is added per 
second), but not during the process of network transmission. Therefore, it is 
necessary to add the aging time before the transmission. Set and adjust this 
parameter according to the actual situation in the low-speed network.
Perform the following configurations in interface view.
Ta b l e 497   Specify Transmit-delay
By default, the time for transmit-delay is 1 second.
Configuring a Stubby 
Area and a Totally 
Stubby AreaUsually, OSPF has 5 kinds of LSA packets, as follows:
■Router-LSA: Generated by each router and transmitted to the whole area, 
describing link status and cost of the router.
■Network-LSA: Generated by the DR and transmitted to the whole area, 
describing the link status of local network segment.
■Net-Summary-LSA: Generated by the ABR and transmitted to relevant areas, 
describing routing of certain network segment of the area.
■Asbr-Summary-LSA: Generated by the ABR and transmitted to relevant area, 
describing routing to ASBR.
■AS-External-LSA: Generated by the ASBR and transmitted to the whole AS 
(excluding the Stub area), describing routing to AS external.
A “stub area” is the area that does not advertise the received external LSA, inside 
which the scale of the routing table and the quantity of the transmitted routing 
OperationCommand
Configure the interval of LSA 
retransmission for the neighboring routersospf timer retransmit seconds
Return to the default value of 
re-transmitting intervalundo ospf timer retransmit
OperationCommand
Set the delay time of LSA tramsmissionospf trans-delay seconds
Return the default value of transmit-delayundo ospf trans-delay 
						

							432CHAPTER 29: CONFIGURING OSPF
information is reduced greatly. A default routing (0.0.0.0) is generated for the area 
by the ABR of the area to insure that these routes are reachable. A stub area is an 
optional configured attribute, but it does not mean that each area is configurable. 
Usually, a stub area is located at the boundary of the AS. A non-backbone area 
with only one ABR or multi-ABR that are not virtually connected between ABRs 
can be configured as a stub area.
A “totally stubby area” is the area that does not receive Type-3, Type-4, and Type-5 
LSA (excluding Type-3 LSA which contains default routing and is generated by the 
ABR). Inside such areas, there is no route to the outside and other areas of the AS, 
so the scale of the routing table and the quantity of the transmitted routing 
information is less. 
A totally stubby area is also an optional configured attribute with the 
configuration conditions that are the same as those of stub areas.
When a stubby area or totally stubby area is configured, the following must be 
noted:
■The backbone area cannot be configured as a stubby area or totally stub area 
and the virtual connection cannot pass through a stubby area or totally stub 
area.
■If one area is configured as stubby area or totally stub area, all routers in this 
area must be configured with this attribute.
■An ASBR cannot be inside a stubby area or a totally stub area, which means 
that the exterior route of the AS cannot be transferred to the area.
Perform the following configuration under OSPF view:
Ta b l e 498   Configure Totally Stubby Area of OSPF
By default, no stubby area or totally stub area is configured. The cost of the 
default routing sent to Stub area is 1.
The area is configured of totally stub area when no-summary option is selected
Configuring an NSSA 
AreaNSSA areas are areas that can import external routing by itself and advertise in the 
AS, but cannot accept external routing generated by another area in the AS. 
Actually an NSSA area is one form of a stub area, which can conditionally import 
AS external routing. A new area-NSSA Area and a new LSA-NSSA LSA (or called 
Type-7 LSA) are added in the RFC1587 OSPF NSSA Option.
The NSSA and stub area are similar in many ways. Neither of them generates or 
redistributes an AS-External-LSA (namely Type-5 LSA), and both of them can 
generate and import a Type-7 LSA. Type-7 LSA is generated by the ASBR in an 
NSSA area, which can only advertise in an NSSA area. When a Type-7 LSA reaches 
the ABR of an NSSA, the ABR selects whether to transform the Type-7 LSA into an 
AS-External-LSA to advertise to other areas.
OperationCommand
Define an area as stub area or totally stub 
area and specify cost value.stub cost cost area area-id [ 
no-summary ]
Cancel Stub Areaundo stub cost cost area area-id [ 
no-summary ] 
						

							Configuring OSPF433
In the following group network, an AS operating the OSPF protocol includes three 
areas, area 1, area 2, and area 0. Area 0 is the backbone area. The other ASs 
operate RIP. Area 1 is defined as an NSSA area. After an RIP route advertises to the 
NSSA ASBR that generates a Type-7 LSA and propagates in Area 1. After the 
Type-7 LSA reaches the NSSA ABR, it is transformed into a Type-5 LSA that is 
advertised to Area 0 and Area 2. The RIP route is generated as a Type-5 LSA and 
propagated in the OSPF AS by the ASBR of Area 2. This Type-5 LSA will not reach 
Area 1 because Area 1 is an NSSA area. On this point, an NSSA area and a stub 
area are the same.
Figure 146   Schematic Diagram of an NSSA Area
Perform the following configuration in OSPF view:
Ta b l e 499   Configure an NSSA Area of OSPF
By default, no area is configured as an NSSA area.
The default-route-advertise attribute generates default a Type-7 LSA. A 
Type-7 LSA default route is generated on the ABR no matter whether there is a 
default route 0.0.0.0 in the routing table while applying this parameter. A Type-7 
LSA default route is only generated on an ASBR when there is a default route 
0.0.0.0 in the routing table.
The no-import-route attribute is used on the ASBR, which allows the OSPF route 
that is imported using the 
import-route command, to not be advertised to the 
NSSA area. If the NSSA router is both ASBR and ABR, this parameter option is 
always selected.
Configuring Route 
Summarization within 
the OSPF Domain
.
Route summary provides that the routing information is processed in the ABR. 
Only one route is sent to other areas for the network segment configured with 
summary. One area can be configured with multiple summary network segments 
so that OSPF can summarize multiple network segments. When the ABR sends 
routing information to other areas, Sum_net_Lsa (Type 3 LSA) is generated for 
each network segment. If there are some continuous network segments in the 
area, they can be summarized into one network segment with a range command. 
Then the ABR only sends one summery LSA and all other LSAs in the summary 
network segment range specified with this command are not sent separately, 
OperationCommand
Configure an area as NSSA areanssa area area-id [ 
default-route-advertise ] [ 
no-import-route ] [ no-summary ]
Cancel the configured NSSA areaundo nssa area area-id
area 0
area 1
NSSA
Router
area 2
Router RouterRIP
Router Router
RIPNSSA
ASBR NSSA
ABR 
						

							434CHAPTER 29: CONFIGURING OSPF
which reduces the LSDB in other areas. The configuration of range is only effective 
when it is configured on teh ABR in the stub area.
For example, there are two network segments in an area as follows:
202.38.160.0    255.255.255.0
202.38.180.0    255.255.255.0
They are summarized into one network segment: 202.38.0.0  255.255.0.0
When the summary network segment of a specific network is added to an area, 
the internal routes with the IP addresses that fall in this summary network 
segment are not broadcast separately to other areas. Only the abstract 
information of the route of the whole summary network segment is broadcast. If 
the network segment range is restricted with the 
notadvertise attribute, the 
abstract information to this network segment route is not broadcast. This network 
segment is described in the form of an IP address/mask. Receiving the summary 
network segment and the restriction of the network segment can reduce the 
inter-area routing information.
Note that the route summary is only effective when configured on an ABR.
Perform the following configurations in OSPF view.
Ta b l e 500   Configure Route Summarization Within OSPF Domain.
By default, inter-area routes are not summarized.
It must be noted that a routing summary configuration is only valid on the ABR.
Creating and 
Configuring a Virtual 
LinkAfter the OSPF area division, all the areas may not be of equal size. One particular 
area is unique and that is the backbone area with the area-id of 0.0.0.0. OSPF 
route update between non-backbone areas is carried out through the backbone 
area. The OSPF protocol requires that all non-backbone areas be connected to 
backbone areas and at least one port on an ABR must be in the area 0.0.0.0. If 
there is no physical connection between an area and the backbone area 0.0.0.0, a 
virtual link must be created.
If a physical connection is not possible due to the limitation of the network 
topology, a virtual link can satisfy this requirement. Virtual link refers to a logical 
connection channel between two ABRs that is created through an area of 
non-backbone area internal routes. Both ends of the virtual link must be ABRs and 
both ends must be configured at the same time so that the virtual link can take 
effect. A virtual link is flagged with the ID of the opposite router. The area 
providing the non-backbone internal route for both ends of the virtual connection 
is called a transit area, whose area-id must also be specified. 
OperationCommand
Configure route summarization within 
OSPF domain.abr-summary address mask mask area 
area-id [ advertise | notadvertise ]
Cancel route summary between areasundo abr-summary address mask mask 
area area-id 
						

							Configuring OSPF435
The virtual link is activated after the route through the transit area is calculated. It 
is equivalent to a point-to-point connection between two terminals. Parameters 
can be configured for this connection like a physical interface, such as sending a 
hello-timer. 
A “logic channel” is provided for multiple routers running OSPF that forwards 
messages between two ABRs. Since the destination addresses of the protocol 
messages are not these ABRs, the messages are transparent to them and they are 
transmitted as ordinary IP messages, while routing information is transmitted 
directly between the two ABRs. Routing information here means an LSA of Type3 
that is generated by ABR. The synchronization of routers in the area is not 
changed.
When configuring a backbone area, note that:
■The backbone area is responsible for advertising the routing information of the 
non-backbone area. If the AS is divided into more than one area, one area 
must be the backbone area, and other areas must be connected with the 
backbone area directly or logically.
■The backbone area must include all ABRs, and may include routers belonging 
to the backbone area only. An ASBR may not be inside the backbone area.
■ABRs inside the backbone area must be well connected, and may be connected 
physically or logically (establishing virtual connection between ABRs).
When configuring a virtual connection, note that:
■A virtual connection can only span one area, which means that the 
non-backbone area can establish virtual connection with the backbone area 
only by spanning one other non-backbone area.
■Multiple virtual connections can be connected in series to form a new virtual 
connection.
Perform the following configurations in OSPF view.
Ta b l e 501   Create and Configuring a Virtual Link
By default, there is no virtual link is created. The attributes for this command have 
the following default values:
■area-id : None
■router-id : None 
■hello-timer: 10 seconds
■retransmit-timer: 5 seconds
■transit-delay: 1 second
OperationCommand
Create and configure a virtual linkvlink peer-id router-id transit-area 
area-id [ hello-timer seconds ] [ 
retransmit-timer seconds ] [ 
transit-delay seconds ] [ dead-timer 
seconds ]
Delete the specified virtual linkundo vlink peer-id router-id 
transit-area area-id 
						

							436CHAPTER 29: CONFIGURING OSPF
■dead-timer: 40 seconds
Configuring 
AuthenticationOSPF supports simple text authentication and MD5 authentication between 
adjacent routers.
Perform the following configurations in interface view.
Ta b l e 502   Configure Authentication
By default, the interface does not authenticate OSPF packets. 
The maximum length of a password for plain text authentication is 8 characters 
and for a MD5 string authentication the maximum length of the password is 16 
characters. The 
key-id attribute is the key value of MD5 authentication, ranging 
from 1 to 255.
Note that the configured packet authentication mode, authentication password, 
and the key-id on the router interface in the same network segment must be 
consistent.
Configuring Route 
Import for OSPFThe dynamic routing protocols on the routers can share routing information. Due 
to OSPF features, the routes found by other routing protocols are always regarded 
as the routes outside the AS in processing. In the receiving command, the cost 
type of the route, cost value, and flag can be specified to overlap default routing 
parameters. 
OSPF uses 4 different route types, whose sequence runs:
■Intra-area route — The route in an area of the AS.
■Inter-area route — The route between different areas of the AS.
■External router Type 1— The received IGP route (such as RIP, STATIC). The 
reliability of this route is high, so the calculated cost of the external route and 
the cost of the route inside the AS are in the same numeric level. It is 
comparable with the cost of OSPF route, i.e. the cost value of external route 
Type 1 = the cost value from the local router to the corresponding ASBR + the 
cost value from ASBR to the destination address of the route.
■External router Type 2 — The received EGP route. Due to the concerns of poor 
reliability of this route, the OSPF protocol considers the cost from the ASBR to 
outside the AS as much as, or more than, the cost to the ASBR within the AS. 
Therefore, mainly the former is considered in the calculation of route cost, i.e. 
the cost value to the external route Type 2 = the cost values from the ASBR to 
the route destination address. If the values are equal, consider the cost value 
from the local router to the corresponding ASBR.
Perform the following configurations in OSPF view.
OperationCommand
Specify a password for OSPF simple text 
authenticationospf authentication-mode simple 
password
Specify the string and key-id for OSPF 
MD5 authenticationospf authentication-mode md5 string 
key-id
Cancel authentication on the interfaceundo ospf authentication-mode