Hello.
We've been discussing an issue which happened to me and my team and thinking of few technical things about it.
It's rather simple - due to routing loop (thanks to static routes which noone really cares about, usually it's just 'there' and 'I better leave it as it is, if it's there then probably there's some reason for this...') we resulted in TONS of ttl-expired icmp packets towards source of the initial packet. And, funny thing, the flow of this ICMP TTL expired packets knocked out a small ASA5505. This is how we became aware of it. It's sorta testing environment and there was no graphs or some other sort of statistics that could indicate us this issue with routing. And we decided to discuss - is there any CLI command which could give some output and show raised amount of ICMP packets? ip accounting? Or it's too severe for CPU?
Any ideas how to detect a routing loop with CLI without taking routing table in bolts and nuts?
PS: cisco/juniper - no big difference, both are interesting
Thanks a lot for your ideas!
No comments:
Post a Comment