BGP peers send a keepalive to each other every 60 seconds by default.
If you are experiencing severe congestion issues, I suppose it's
possible that those keepalives are being missed. If so, that would
cause the BGP session to drop and then come back up. This *really*
creates more congestion problems if you're passing a lot of routes
because the peers have to pass their entire tables every single time the
line flaps. If you're having bandwidth issues this will make things
much worse.
I'd first try to address the traffic issue. Is all of this traffic
necessary or can you filter out some of it to relieve congestion?
If it's all necessary, is some higher priority? Can you configure
local policy routing to make BGP information high priority? You can set
its IP precedence to critical or higher and that might help.
You might also try to change your queueing mechanism. Is this a T-1 or
lower? Perhaps try some form of WFQ or CBWFQ. If it's faster than T-1
make sure you're using FIFO as your queuing mechanism.
Are you fast-switching traffic from a high-speed interface to a
low-speed interface? If so, if you have the CPU available, turn off
fast switching. That might slow things down a bit.
You actually have a lot of options depending on the specific cause of
the problem and your desired end result.
HTH a little,
John
>>> "KroyweN" 8/29/01 7:23:39 AM >>>
Gentlemen;
I am experiencing the flapping of our bgp everytime it hits its
maximum
bandwidth . The line was BER tested ok, im using 7206 router. do u have
any
idea why it flaps is it on the router configuration?
thank you,
kroywen
Message Posted at:
http://www.groupstudy.com/form/read.php?f=7&i=17685&t=17664
--------------------------------------------------
FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html
Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
