It takes too long when Hello packet is lost. Repairing split flooding topology 
needs to be fast.


Fortunately, lost hello packets are a relatively rare occurrence.  While 
repairing the flooding topology needs to be done expediently, attempting to do 
so and triggering a cascade failure of the network is counter-productive. Given 
this alternative, a bit of extra delay when adding a new system to the network, 
or trying to recover from multiple failures seems wise. Rushing and making 
things worse does not.  The first
priority must remain network stability.



[Les:] Just to put this minor issue to rest…
It would also be possible for the node making the temporary flooding request to 
send hellos at a faster rate (say once a second rather than once every 10 
seconds) until it sees that the neighbor is sending LSP/SNP updates.
This just isn’t a significant issue.

   Les


_______________________________________________
Lsr mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/lsr

Reply via email to