OSPF :
OSPF is link state IGP protocol
OSPF uses IP protocol 89
OSPF uses a shorted path first algorithm in order to build data path and ... The shortest path is calculated with the use of the Dijkstra algorithm.
OSPF send trigger update if there is change in network
OSPF send periodic update of entire routing table every 30 minute
in OSPF all area must connect to area 0
all router in an area have the same topology table
all area must have hierarchical ip addressing design , so that summarization with in the area would be easier.
in OSPF ABR and ASBR can only do summarization
Rule : OSPF router id
all ip address configured with in interfaces in a router , the highest ip will elect as ospf router id. however loop back interface will beat physical interface , even loop back has a lowest ip address .if there is multiple loop back , highest loop back gets elected .OSPF router id keep changing if a router or process gets reboot .always better to hard code router id .
OSPF neighbor states :
OSPF down state : if router send out a hello message on a interface still did not hear back , this is called down state
OSPF init state : if router receive a hello message ( receive hello) , the router goes to init state .
OSPF 2 way state : when a router reply to hello message , router goes to 2way state .
OSPF ex-start state : if hellos and all are agreed it comes to master and slave selection process , where router goes to ex-start state . the master sends DBD packet ( database description packet ) and then slaves sends its DBD back to master .
OSPF Loading state : once DBDs received and acknowledged router goes to loading state .
OSPF full state : once neighbor synchronized with database router goes to full state .
hello message are send out once very 10 second in broadcast/point2point network , and every 30 second in no broadcast network
dead time is 4 times to hello timer
to form a neighbor OSPF router must match hello and dead timer , network mask , area id and authentication .
OSPF maintains 3 table :
neighbor table :
topology table : knows about network with in a area
routing table :
on a OSPF broadcast segment, OSPF DR/BDR gets elected
all OSPF speaking router communicate on multicast address 224.0.0.5
DR/BDR routers communicate on 224.0.0.6
OSPF packet type :
hello
Database description DBD
LSR : link state request
LSA : link state advertisement
LSU : link state update
LSACK : link state acknowledgement
OSPF network type :
broadcast multi-access
non-broadcast multi-access default on serial interface
point to multipoint RFC standard advertise /32 for each link
point-2-point Cisco proprietary default mode on sub interface
broadcast
point 2 multipoint non broadcast Cisco proprietary static neighbor configuration required
OSPF ABR summarization happen with “ area --- range “ command , and it creates a summary route or null 0 route in local router .
OSPF ASBR summarization happen with “ summary address” command in OSPF process .
OSPF LSA type :
LSA type 1 : router LSA (routing update )*****
LSA type 2 : network LSA (DR generated )
LSA type 3 : summary LSA ( ABR summary route) route from ABR
LSA type 4 : summary LSA ( ASBR location ) ip address of ASBR advertise
LSA type 5 : External LSA (ASBR summary route )
Some OSPF handy command :
show ip ospf nei
show ip ospf int
clear ip ospf process
show ip ospf database
LSA (link state advertisement ) 1 -- originated by every router in an area
In OSPF broadcast segment hellos are send as multicast
In OSPF non broadcast segment hellos are send as unicast
on OSPF non-broadcast network , neighbor has to be configured on
OSPF DR router manually by neighbor command.
In general loopback interface route will be treated as host route in OSPF
if loopback need to advertise as a network , use ip ospf network point-to-point
command on loopback interface .
for OSPF network type point to point hello time 10 dead time 40
for OSPF network type multipoint hello time 30 dead time 120
point to multipoint network does not have DR/BDR election send hellos as multicast to
224.0.0.5, its change the next hop value .
inter area route is LSA 3 shows as OIA in routing table
intra area routes are LSA 1 & 2 shows as O in routing table
100
OSPF cost = -----------------------
Interface bandwidth in Mbps
every shared network should have DR and a BDR
DR BDR election :
Highest priority on interface is better
if it is tie then highest OSPF router ID better
OSPF area :
stub area : block type 5 lsa entering into this area instead gets a default route from neighbor O*IA mark, along with inter area route, this is industry standard
totally stub area : block type 3 , 4 and 5 LSA entering to this area , this is Cisco proprietary , to enable this use 'area x no-summary’ on border router command
router gets only a default route o*IA.
not-so-stuby area : passes external (type 5) LSA via type 7 LSA to the backbone area 0 and once it passes not so stuby area converted by ABR, the LSAs shows as type 5
meaning external route . router receive a O*IA default route
OSPF does not use key chain method of authentication , it uses authentication on interface , support two type authentication , md5 and clear text .if we use ip OSPF authentication-key command , we are using clear text authentication.
OSPF routes type :
E1 : cost of route increments as it passes through the autonomous system .
E2 (default) : cost of route remain same as it passes through autonomous system this is default type for OSPF external route
use auto-cost reference bandwidth to adjust cost as require in OSPF process .
Default-information originate ' command will send a default route to the network o*E2 mark , router should have a static route, if there is no static route use ' default-information originate always ' command to advertise a default route.
thanks alot
ReplyDeleteVery useful and organized data