RE: Traceroute versus other performance measurement
Presumably you're asking if it's a good tool to measure *available* bandwidth or lack thereof, i.e. congestion and its byproducts of packet loss and increased latency. No, it isn't! - Congestion resulting from asymmetric paths can be misinterpreted through traceroute. - Cases where ICMP performance with respect to the routers themselves is significantly lower than throughput of production traffic will often skew results. Having said that, where traceroutes suggest a POSSIBLE problem on my own network, I'd check further. However, I would never ask the operator of another network to troubleshoot solely on the basis of traceroute output.
-----Original Message----- From: Paul Bradford [mailto:paul@adelphia.net] Sent: Wednesday, November 29, 2000 9:08 AM To: nanog@merit.edu Subject: Traceroute versus other performance measurement
I need help with a reality/sanity check. Traceroute is a good tool for checking for routing type problems (loops). Does anyone feel it's a good tool to use for testing "bandwidth"....
Mark Borchers wrote:
Presumably you're asking if it's a good tool to measure *available* bandwidth or lack thereof, i.e. congestion and its byproducts of packet loss and increased latency.
No, it isn't!
- Congestion resulting from asymmetric paths can be misinterpreted through traceroute.
- Cases where ICMP performance with respect to the routers themselves is significantly lower than throughput of production traffic will often skew results.
Having said that, where traceroutes suggest a POSSIBLE problem on my own network, I'd check further. However, I would never ask the operator of another network to troubleshoot solely on the basis of traceroute output.
-----Original Message----- From: Paul Bradford [mailto:paul@adelphia.net] Sent: Wednesday, November 29, 2000 9:08 AM To: nanog@merit.edu Subject: Traceroute versus other performance measurement
I need help with a reality/sanity check. Traceroute is a good tool for checking for routing type problems (loops). Does anyone feel it's a good tool to use for testing "bandwidth"....
Also, on some routers, traceroute requires going through the "slow path" (i.e., the router CPU), and show delays much larger than actual operational packets will encounter. -- Regards Marshall Eubanks T.M. Eubanks Multicast Technologies, Inc 10301 Democracy Lane, Suite 410 Fairfax, Virginia 22030 Phone : 703-293-9624 Fax : 703-293-9609 e-mail : tme@on-the-i.com tme@multicasttech.com http://www.on-the-i.com http://www.buzzwaves.com
participants (2)
-
Mark Borchers -
Marshall Eubanks