> The cure for this is customer education, not making the router more > vulnerable to DDoS attacks, either deliberate or unintentional ones. > > ;>
Agreed, but the CPU scheduler should still not allow what in QoS designs we call "default queue starvation". A bulk priority task like a massive BGP update which takes tens of seconds to complete is not expected to be realtime anyway, so it should not stall small low/default priority tasks. We are doing this in QoS, and we are doing this in PC/server OS'es, so why don't we do it in our Router OS? Perhaps intelligent reservation and queuing of resources is not what our vendors are voluntarily pushing, because "just buy bigger boxes" is such a good answer for them, even though its not a solution at all. Regards, Lukas _______________________________________________ cisco-nsp mailing list [email protected] https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/
