0

I'm load testing some web services hosted on IIS7. The web services make network calls to underlying services in response to client requests. The client requests are simple HTTP requests but the requests that emanate from the web service layer start with a ping request (ICMP echo). After almost 6-7 days I have noticed multiple ping request failures and when I monitored the traffic on Wireshark I was able to see that these co-related to multiple ICMP Destination unreachable errors. Intermittently some successful ICMP echo are also present.

Can anyone suggest a good starting point for my investigation? The frequency of hits on the server are ~200 http requests/sec.

2 Answers 2

0

If the ping fails intermittently, you may have an issue with the network adapter, the router or gateway device, the cabling, or other connectivity hardware.

1
  • The ping failure happens only after 6-7 days. Once I restart the server and then restart the load test this behavior is repeatable.
    – Krishter
    Feb 8, 2013 at 18:50
0

Assuming that this isn't a flaky cable (which I've come across as the most common root of these types of problems before, but not the only issue), the first thing that I would do would do is to do a tracert from the client to the server to understand all the points of the network where the ICMP request must traverse. Essentially the reason for this is to correlate all of the IP addresses of all of the network hops and remove them from the equation one by one. Network card, network cable, switch, router, etc.) On the server side, running some type event monitoring on the server to track the networking and % processor idle performance during the time of the Destination unreachable errors would help you get an idea if the network card or server had a process that was overwhelming the server's capacity to accept the request. I assume you're on a gigabit or higher network?

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .