03-02-2010 09:18 PM - edited 03-01-2019 02:17 PM
I have a number of 6500s with SUP720-3BXL and 1GIG of memory. Taking full tables form only one ISP (310K routes) I see CPU utilization as high as 100% for brief periods of time on all of these routers. BGP Scanner is always at the top and it's scan rate of every one minute seems to match the rise and fall of CPU utilization. Question: does anyone else see this on their 6500s? I am running 12.2(14) or so code and doing nothing fancy on this router (even cut off Netflow). I have a dozen or so of the exact routers around the country all with intermittant high CPU utilization. Seems very hard for me to believe that even with full BGP tables I should see CPU go to 100%, even if it for a brief period of time. It may be time for a Juniper...
03-03-2010 02:43 AM
Hello Tedd,
I would suggest you to post the BGP configuration and a show module to see what type of DFCs are on the linecards (if any) we had an issue in which by error a linecard was changed and the new one hadn't the right DFC type, and this caused cpu to go to 100% (it couldn't handle the whole CEF table with 300000 prefixes)
We have two 6500 with sup720 3BXL multihomed with two ISPs each and 12.2(33)SRC and they are working fine.
Hope to help
Giuseppe
03-03-2010 09:07 AM
BGP config is very standard.
Mod Ports Card Type Model Serial No.
---
03-03-2010 10:31 AM
I have a number of 6500s with SUP720-3BXL and 1GIG of memory. Taking full tables form only one ISP (310K routes) I see CPU utilization as high as 100% for brief periods of time on all of these routers. BGP Scanner is always at the top and it's scan rate of every one minute seems to match the rise and fall of CPU utilization. Question: does anyone else see this on their 6500s? I am running 12.2(14) or so code and doing nothing fancy on this router (even cut off Netflow). I have a dozen or so of the exact routers around the country all with intermittant high CPU utilization. Seems very hard for me to believe that even with full BGP tables I should see CPU go to 100%, even if it for a brief period of time. It may be time for a Juniper...
Hi,
The BGP Scanner process walks the BGP table and confirms reachability of the next hops. This process also checks conditional advertisement in order to determine whether BGP should advertise condition prefixes and/or perform route dampening. By default, the process scans every 60 seconds.
The following taskare performed by BGP scanner:-
A check of the next hop that is referenced in the router BGP table
Verification that the next-hop devices can be reached
Thus, a large BGP table takes an equivalently large amount of time to be walked and validated. The BGP Scanner process walks the BGP table in order to update any data structures and walks the routing table for route redistribution purposes. Both tables are stored separately in the router memory. Both tables can be very large and, thus, consume CPU cycles.
Check out the below link for troubleshooting high CPU due to BGP scanner process hope that help !!
http://www.cisco.com/en/US/tech/tk365/technologies_tech_note09186a00809d16f0.shtml#scanner
Remember to rate the helpful post !!
Ganesh.H
03-03-2010 10:38 AM
I think we all know what the role of BGP Scanner is. I really don't need any tired years old Cisco link. I need to hear from folks who are in large ecommerce/SP positions that actually have similar setup with mulitple ISPs and taking full route from all of them.
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide