[quagga-dev 10317] Re: [PATCH] ospfv3-make-maxage-thread-periodic.patch

Christian Franke chris at opensourcerouting.org
Mon Mar 4 17:11:46 GMT 2013


On 03/04/2013 04:01 PM, Dinesh Dutt wrote:
> On 03/04/2013 03:44 AM, Christian Franke wrote:
>> One thing I am still a little bit uncertain about: is it acceptable for
>> a MaxAge LSA to sit around up to 60 seconds in the LSDB while it is
>> already expired? (The specification of the aging mechanism for the OSPF
> This is how OSPFv2 behaves, AFAICT. MaxAge LSAs are not considered hits
> when searching for LSAs (though there were a couple of places where the
> check did not happen).

Yes, OSPFv2 behaves somewhat similar. In difference to OSPFv3 is also
seems not to have per LSA expiry timers set up but does periodic sweeps
of the LSDB. (Which I guess had been a performance dicated decision as
the timers in thread.c are kept in a single linked list which probably
won't scale well if there is a larger number of LSAs each with its own
expiry timer active)

This behaviour however causes an ANVL OSPF compliance test (24.6) to
fail. Therefore, I was wondering whether it is acceptable, however it
may as well be that the test case is just a bit too pedantic there.

-Christian

[1] http://jira.isc.org:8090/display/osr/ANVL+OSPFv2+Compliance+Test+Plan




More information about the Quagga-dev mailing list