View source for Plugga CCNP Route--OSPF

Main Page [http://hermes.cnap.hv.se/wiki/index.php/Plugga_CCNP_Route CCNP Route plugg]
= OSPF =
==States==
* Down -- No hello recieved, but sent
* (attempt) only neghbor command on nmbma
* Init -- Hello recieved, but my Router-ID was not in the neighbor list
* 2-way -- Go here if HELLO or DBD recieved (Dflt hello timer=10 on broadcast & p2p, 30secs in NBMA & Multipoint
*: Don't come here if: 
*:# Network-mask (exceptions exist). Network-number not checked ?!?
*:# Area ID 
*:# Hello-timer and Dead-timer 
*:# Password  
*:# CCIE: Stub flag of network must match (usually also "link-type", but not always )
*: at the end of this stage: do DR/BDR election
*: Here: Reset the Dead-timer (Default is = 4xHello-timer)
* EXstart -- Exchange DBD (a summary = All LSA headers)
*: The router with the higher router ID becomes the master and starts the LSU exchange (includes LSA headers only), and as such, is the only router that can increment the sequence number.
*: If Priority is high and Router-ID is low, the DR might be the slave !!
* EXchange
*: Slave sends LSR to fill in the gaps of the LSU(to populate the LSDB with full LSA's). Master sends LSU with full LSA's. THEN Master->Slave
*: LSU's are ACKnowledged with LSACK-packet
* Full state -- Normal state (2-way is normal with two non-DR/BDR are neighbors)

Good explanation: [http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a0080093f0e.shtml cisco.com-theory]

== Commands ==
  interfa XX 0/0
     ip ospf cost <1-65535>
     ip ospf authentication-key KALLE (remember router conf)
     ip ospf network point-to-point
     ip ospf priority 255  (0=do not participate)
  
  Router ospf PID
    router-id 1.1.1.1
    ospf auto-cost reference-bandwidth <1-65535>
    area x authentication    (remember intf conf)
    redistribute static subnets metric 200 metric-type 2
    network 10.2.2.1 0.0.0.0 area 0             
    network 192.168.1.0 0.0.0.255 area 10       
    area 10 range 192.168.0.0 255.255.0  (ABR only)
    summary-address 172.16.0.0 255.255.0.0  (summerize IN-redistributed routes)
  -------
  ip ospf message-digest-key 1 md5 kal  
  area 0 authentication message-digest

== OSPF Multicast ==
224.0.0.5 (FF02::5) to all others
224.0.0.6 (FF02::6) to DR and BDR
== Modes (chaos in writing) ==
<br/>1. Non-broadcast (NBMA) -- RFC Standard -- neighbors manually configured, DR req'rs full connectivity
<br/>2. Point.to.multipoint -- RFC Standard -- no manual neighbors BUT pseudo broadcasts req'rd (f-r map 103 1.2.3.4 '''broadcast''')
<br/>3. point-to-point -- cisco proprietary -- PPP
<br/>4. broadcast -- cisco proprietary  --- ETHERNET
<br/>5. point-to-multipoint, non-broadcast -- cisco propr. --no pseudo broadcasts req'rd BUT manual neighbors  

* interface serial 0/0.1 point-to-point defaults OSPF network typ to POINT_TO_POINT
* interface serial 0/0.2 '''multipoint''' defaults OSPF network typ to '''NON_BROADCAST'''

The page [http://blog.ipexpert.com/2009/08/13/ospf-over-frame-relay-part-5-point-to-multipoint-nonbroadcast/ ipexpert.com] mentions:
* <tt>ip ospf network point-to-multipoint non-broadcast</tt>

<br/>1. use <tt>frame-relay map ip 1.2.3.4 102 '''broadcast'''</tt> and <tt> neighbor</tt> command
<br/>2. <tt>interface serial 0/0.1 point-to-multipoint<br/></tt>
<br/>&nbsp; <tt>frame-relay map ip 1.2.3.4 102 '''broadcast'''</tt>
<br/>3. <tt>interface serial 0/0.1 point-to-point</tt>

== different types form neighbours ==
  Hub-router# interface serial 0/1.1 multipoint
                ip ospf network point-to-multipoint            show ip ospf interface reveals POINT_TO_MULTIPOINT, Hello 30, Dead 120
   
  Spoke5# interface serial 0/1.1 point-to-point                sho ip ospf int '''initially''' reveals POINT_TO_POINT, Hello 10, Dead 40
          ! Now change the timers
          !
Hub-router# interface serial 0/1.1 multipoint
                ip ospf hello-interval 10                      show ip ospf interf '''now''' reveals POINT_TO_MULTIPOINT, Hello 10, Dead 40
                                                               AND the neighbors form (even though different network types)
== Virtual-link ==
On Abr between area1 and area2 (other end (2.2.2.2) connects to area0
 R3# router ospf 1
    router-id 3.3.3.3
    area 1 <through area> virtual-link 2.2.2.2 
 R3# router ospf 1
    router-id 2.2.2.2
    area 1 <through area> virtual-link 3.3.3.3
== NEW: Show ip ospf BORDER-ROUTERS ==


  Router# '''show ip ospf border-routers'''
  
  OSPF Process 109 internal Routing Table
  
  Codes:  i - Intra-area route, I - Inter-area route
  i 192.168.97.53   [10] via 172.16.1.53,   Serial0, ABR,  Area 0.0.0.3, SPF 3
  i 192.168.103.51  [10] via 192.168.96.51, Serial0, ABR,  Area 0.0.0.3, SPF 3
  I 192.168.103.52  [22] via 192.168.96.51, Serial0, ASBR, Area 0.0.0.3, SPF 3
  I 192.168.103.52  [22] via 172.16.1.53,   Serial0, ASBR, Area 0.0.0.3, SPF 3

== Hard Questions/Answers ==
#How to form neighbors on Frame-Relay
#:A) ip ospf network broadcast
#:B) ip ospf network point-to-point
#:C) neighbor 10.255.255.1  (IP, not ID) of neighbor closest interface)
#Which three restrictions apply to stub areas
#: Stub area can not be Backbone
#: No ASBR are allowed
#: "No virtual links are allowed"
#In which state does DR and BDR establish adjecency with each OSPF router in the network?
#: Init State
#The maximum number of routers per OSPF area typically depends on which three factors?
#: The kind of OSPF areas being implemented
#: the number of external LSAs in the network
#: how well the areas can be summarized
#Which three are advantages to create multiple areas in OSPF
#: Less frequent SPF calculations (Rob: changes create recalc's)
#: smaller routing tables (Rob: ABR can summarize)
#: reduced LSU overhead (Rob: ABR can summarize)
#How to create a stib-area of area 3
#:on ABR, area 3 stub no-summary
#: (Rob: it is better to answer one correct then three half-bad, but working ie "area 3 stub no-summary" in all routers in area 3
#What to check first when problem with OSPF virtual links
#: OSPF Router ID (Becouse "area <area> virtual-link <'''router-id'''>
#R2 stuck in INIT -state ??
#: R2 has not recieved the OSPF hello from R1 router
#: (: INI-state == Sent HELLO, not (yet) got any HELLO back)
#When learning a new route, if a LSA recieved is not found in the topology database, what will an internal OSPF router do?
#:The LSA is flooded immediately out of all the OSPF,interfaces, except the interface from which the LSA was recieved.
# Exhibit
#: <tt>O IA 6.0.0.0/8 [110/65] via 5.0.0.2, 00:00:18 Serial 2/1/0</tt>
#: <tt>O*N2 0.0.0.0/0 [110/1] via 5.0.0.2, 00:00:18 Serial 2/1/0</tt>
#: "CISCO"  (;-) ANSWER: OSPF Router 5.0.0.2 is an ABR
#LSA flapping from source "172.16.20.2" of LSA for 172.16.10.0
#: LSA Content seen from "show ip ospf database" changes repeatedly, everey few seconds
#: ANSWER: Two routers have the same OSPF router ID
#What are two IOS commands that can be used to view neighbor adjacencieds
#: show ip ospf neighbors
#: show ip ospf interfaces (Example: ... Adjacent neighbor count is 1)
#Exhibit! Fixing OSPF without fixing links
#: D1:area 16 virtual-link 8.187.175.82 (Loopback of D2)  
#Highest Priority/RouterID/Loopback/IP wins DR/BDR Election
#ip ospf network nonbroadcast
#: True DR/BDR election will occur (due to NMBMA)
#: "neighbor" command is required (No multicast(broadcast) HELLO's exists)
# show ...current state of OSPF database loading process.
#: show ip ospf neighbors
#External routes are not showing up. What to check first?
#: subnet-keyword in redistribution?
#: Are the ASBRs trying to redistribute the external routes into a totally stubby area?
== NBMA Table ==
* non-broadcast (RFC-compliant)
**One IP subnet. 
**Neighbors must be '''manually''' configured. 
**DR and BDR are '''elected'''. 
**DR and BDR need to have full connectivity with all other routers. 
**Typically used in a full- or partial-mesh topology.

*point-to-multipoint (RFC-compliant) 
**One IP subnet. 
**Uses a multicast OSPF hello packet to automatically '''discover''' the neighbors. 
**DR and BDR are not required. The router sends additional LSAs with more information about neighboring routers. 
**Typically used in a partial-mesh or star topology.

*point-to-multipoint nonbroadcast (Cisco proprietary) 
**If multicast and broadcast are not enabled on the VCs, the RFC-compliant point-to-multipoint mode cannot be used, because the router cannot dynamically discover its neighboring routers using the hello multicast packets; this Cisco mode should be used instead. 
**Neighbors must be '''manually''' configured. 
**DR and BDR election is not required.

* broadcast (Cisco proprietary) 
**Makes the WAN interface appear to be a LAN. 
**One IP subnet. 
**Uses a multicast OSPF hello packet to automatically '''discover''' the neighbors. 
**DR and BDR are '''elected'''. 
**Full- or partial-mesh topology.

*point-to-point (Cisco proprietary) 
**Different IP subnet on each subinterface. 
**No DR or BDR election. 
**Used when only two routers need to form an adjacency on a pair of interfaces. 
**Interfaces can be either LAN or WAN.
** Rob says '''discover'''