Thursday, 16 May 2019

OSPF Series 7: Example of OSPF Path Preferences, Troubleshooting neighborship stuck in various OSPF States

SCENARIO: Path Preferences:

  • Configure Router-IDs on each router

  • COnfigure IP address of interfaces in each router and add them using network command of OSPF

  • Note: there is no need to configure static routes on MT708 as both the links are directly connected. MT708 doesn’t run OSPF.

  • (On MT704, give static route #ip route 172.16.18.0/28 172.16.48.8) No need to give this since we can advertise as directly connected and MT704 will also be an ASBR (also MT701 will be an ASBR)

  • On MT701, give static route #ip route 172.16.48.0/28 172.16.18.8

  • On MT704. #router ospf 1

    • #redistribute connected //only redistribute connected here...don’t redistribute static
  • Now, we can see an E2 route on the other routers

  • img

  • On MT701.

    • #int lo1
    • #ip address 1.1.1.1/32 //we make a loopback and give the network in area 0 so that the router MT701 becomes an Area Border Router
    • #router ospf 1
    • #redistribute static //we need this as we need a common network path
    • #redistribute connected
    • //after this we will get even 172.16.18.0/28 on MT702 as E2
  • To get the entries connected to MT701 externally as E1, on MT701:

    • #route-map a
    • #set metric-type type-1
    • #redistribute connected route-map a
    • #redistribute static route-map a
    • img
  • To get the entries connected to MT704 externally as E2, on MT704:

    • #route-map new
    • #set metric-type type-2
    • //now, the loopback 4.4.4.4 on MT704 will be shown as E2 on MT702’s routing table
  • There is no need for any configurations to get IA. Any network in different area will automatically show up as O-IA. On MT702, since it is in area 0, area 1 and area 3, 172.16.13.0/29 n/w will be inter-area as it is in area 2. So, MT702:

    • img
  • SO→ If MT701 will say the route to IP address as E1 but MT704 will say the same route as E2. But, all routers will store E1 route as it is more preferable. But, if we check the LS Database, we can see both E1 and E2

 


 

TROUBLESHOOTING: OSPF STUCK IN STATES:

http://www.cisco.com/c/en/us/support/docs/ip/open-shortest-path-first-ospf/13699-29.html

#show ip ospf neighbor -> to check stuck

2 WAY STATE Stuck:

  • Priority is 0 for both

To verify this, show ip ospf nei detail (show priority and check if neighbor is telling itself as DRO)

After this, we can use show ip ospf interface brief (and check if we are also advertising as DRO)

EXSTART STATE Stuck:

  • Different MTU on both sides

show ip int br (to check our MTU- do on both sides)

  • Putting ACL for the unicast IP so that the DBDs do not reach

EXCHANGE STATE Stuck:

  • MTU Mismatch itself....The router with higher MTU can go to Exchange and will be stuck in Exchange. The router with lower interface MTU will be stuck in Exstart state

LOADING STATE Stuck:

  • If we use the same IP address on two different VLANs (SVIs participating in OSPF) on different switches....In this case is not possible to get to 'Full' state as there are two different LSAs coming from the same IP Address.
  • Neighbors that do not transition beyond this state are most likely exchanging corrupted LSAs. This problem is usually accompanied by a %OSPF-4-BADLSA console message) Note: this is uncommon issue)

TROUBLESHOOTING:

http://www.cisco.com/c/en/us/support/docs/ip/open-shortest-path-first-ospf/7112-26.html#anc5

 

No comments:

Post a Comment