Showing results for 
Show  only  | Search instead for 
Did you mean: 

Network Latency between collector-agent the influence on system?



Maybe anyone know, the what is happened if Latency between Agent -- Collector very low.

When using the Nginx we have issue when Web-server hangs only using the Dynatrace agent.

(If remove agent - no hangs).

So what is the behavior of instrumented web-server in this case?

ping dynatr-collector:64 bytes from dynatr-collector ( icmp_seq=2 ttl=124 time=179 ms

And nginx has state R, and cpu 100% after a few hours working.

Maybe some ideas?

Thanks, Aleksandr.


From what I understand (Although doesn't fully answer your question),

1. CPU 100% have nothing to do with latency (Suggest to open a support ticket for it)

2. AFAIK, documentation would suggest that the latency should be <10ms, otherwise most PurePath would be corrupted, truncated or timeout. (Anyway, this just affect the monitoring, not the SUD)

Anyway, someone which is more experienced and expert than me should be able to answer your doubt, or, correct my this statement if I am wrong about this.

Hi Wai, How to test latency between these? Any tool?

Based on MY EXPERIENCE, a simple ping test to the collector, on the machine where the agent is installed is already enough.


Hello @Oleksandr M. & @Rajesh S.

May be there will be some other ways to find the latency between agent and collector but if you add the additional columns (below screenshot as a reference) in the 'Agents Overview' dashboard to see the average and maximum instrumentor response time.