""MADMAN"" wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... > Priscilla Oppenheimer wrote: > > > > Well, maybe I overstated it a bit. ;-) My main complaint about the debug > > commands is that the output is too cryptic. Also, some of them were clearly > > designed for the Cisco developers not for the end user of the router > > (network admin, engineer). The information they provide is simply not > > helpful. > > > > Inserting a sniffer can definitely be a pain on a WAN, on the other hand. > > Plus WAN sniffers are terribly expensive. Actually inserting a sniffer is > > more of a pain than it used to be on LANs too. But at least the result is a > > plain-language decode of every packet. > > > > By the way, do you remember which EIGRP debug commands you used and how > > they helped solve the problem? That might be helpful info for us (if you > > have time to explain, no biggie if you don't.) > > > > Thanks > > > > Priscilla > > Actually I used debug eigrp packet found a couple of neighbors were > bouncing eratically which I had also noticed in the ip routing table. I > tried pinging these neighbors and was loosing many packets, this is over > a 100M ethernet. Since this customer mentioned that they had done some > work on a Microsoft server including adding a second interface (arghhh) > I had a good suspect. Since I have seen in the past multiinterfaced > servers do wierd things like foward multicast packets I suspected a > possible routing loop. I enabled debug ip icmp and basically crashed > the MSFC. It was sooooo busy spewing out ICMP TTL expired messages that > caused the CPU to hit 99% and the router was not able to maintain it's > routing functions etc... I asked the customer to grab the server guy > and have him shut down the second interface, problem solved.
CL: you sure you didn't say something more like "grab the server guy and throttle him a good one!" ??? > > The IP ICMP debug was really the helper here but the point is I was > able to find the problem using debug, I'm 300 miles from this customer, > much more quickly than finding someone locally who could drive a sniffer > and read/email the output. I admit crashing the router was not good but > "normally" a ip icmp debug will not do that hence I say use any debug > with some caution and customer warning, this may be hazardous to your > network!! > > Dave > > David Madland > Sr. Network Engineer > CCIE# 2016 > Qwest Communications Int. Inc. > [EMAIL PROTECTED] > 612-664-3367 > > "Emotion should reflect reason not guide it" Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=44993&t=44876 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
