OSPF database question

Unanswered Question
Jun 11th, 2008


I have a bit of a weird issue. I have 4 link-id's in my ospf database with the advertising router being the router where the routes are being learned from. These 2 networks used to be configured on an old leased line out to the remote site but I have since shut them down. Still seeing these networks in though in the OSPF database but not in the routing table as you would expect. My question is is there any way of clearing down the ospf database for these 4 networks without having to reset the whole ospf process on the router. Also, the Seq # is incrementing all the time - why would that be if the wan out to the remote site is now shut?

Any help appreciated?



Test_7604>sh ip ospf database | include 10.72.128

Link id Adv Router Age SEQ # 1388 0x80002826 0x00E59A 1388 0x80002826 0x005CF5 1390 0x80001FC8 0x00BB2B 1390 0x80001FC8 0x003286 1648 0x800000EF 0x00CA14

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
tarunavakonar Wed, 06/11/2008 - 02:00

what i know u can use passieve-interface serial xx command on router ospf process

ashok_boin Wed, 06/11/2008 - 02:56

You may try with "clear ip ospf " which may help.

Or, if it's no more a neighbor, then these will be automatically gets removed after MaxAge.



maryodriscoll Wed, 06/11/2008 - 03:48


Just checked and would you believe they are now gone from the ospf database. What is the maxage timer out of interest or can you point me to a good ospf doc?



ashok_boin Mon, 06/16/2008 - 21:30

Hi Mary,

By default, the max age timer for OSPF is 1 hr.

But, I have seen in your first e-mail that the seq number is increasing still means it's still getting update from same router (router ID).

Can you give "sh ip os data router x.x.x.x" to see more details about this? Please take the above command output 2 times with a gap of 1 hr.

This can be a good resource for OSPF.




Giuseppe Larosa Wed, 06/11/2008 - 04:21


OSPF is a Link state protocol so there can be an entry in the OSPF database with link state =down instead of removing the whole entry from the database as you can expect from a distance vector protocol.

the state = down can be signaled with a infinite metric that for ospf is a 24bit number of all 1s.

As you have noticed the ip networks have been correctly removed from the ip routing table.

You should look at the detail of the ospf database to see what are the fields values for each LSA data structure.

try to use: sh ip ospf database router

or sh ip ospf database adv-router

Hope to help


Giuseppe Larosa Wed, 06/11/2008 - 04:25


there can be a transient time where you see the older LSAs and the newer ones both in the DB with the clear ip ospf process-id you restart the process and you will see the new updated DB.

best regards


merryllem Mon, 06/16/2008 - 23:19

i've encountered a similar scenario where i needed to clear the contents of the ospf database. I tried everything .. clearing process, changing process id.. and even unconfiguring OSPF. The only way I clear the contents of the ospf database was to reload the router.


This Discussion