Subject: psping.exe -q doesnt really make it quiet
Posted: 25 December 2014 at 11:57am
Hmm, I'mnot sure why you are asking about "video jitter" when discussing anetworking issue :)
I'm talking about network jitter, of course.http://en.wikipedia.org/wiki/Jitter
I let the graphs render in PRTG for about an hour (after my previous post), andit's pretty clear PsPing has some serious issues with "fake" jitterin its measurements/results. It is particularvisible on very low latency ping requests. Any host below 10ms seems to begiving highly inaccurate results - with the numbers being pretty muchunusable.
The hosts I am pinging is a local university (www.uio.no) and a local mediahouse (www.ap.no). Both servers are extremely reliable from my location(I am connected one hop away from the NIX/backbone), and they give me resultsof roughly 2 ms. But PsPing is giving me results anywhere ranging from 3 ms up to100 ms.
I am letting PsPing send 5 ping requests, and I grab the average value to use in my graphs. This is done once a minute.
I did a comparison with hrPING (http://www.cfos.de/en/ping/ping.htm), and thistool is giving very different and a lot more reliable results.
hrPINGgives me a nice and steady result of ~3ms when pinging www.uio.no andwww.ap.no. It's off by one millisecond compared to PingPlotter and Windows'internal ping command, but that doesn't really matter as long as it does notproduce fake jitter like PsPing does.
It'srunning on a Core 2 Windows 7 64bit system. There are some spikes in processorusage from running the PRTG server in the background (once every minute), but I don't understandwhy it should affect the PsPing measurements. I tried looking in PerfMon towatch the CPU usage, but I'm unsure if there are correlations between CPU usageand PsPing jitter. Sometimes there are spikes in PsPing's output when therearen't really any CPU use at all.
I've moved away from PsPing and started using hrPING in PRTG instead, and the scripts are all re-written to accommodate the output from hrPING, so unfortunately I'm not able to keep troubleshooting PsPing for you at this time.
PsPingclearly has some big problems with how it's calculating/generating its results,and the developer really should look into it.
As a final note, here is console output from PsPing and hrPINg in respective windows running at the same time (1 second interval)
Here is PingPlotter output from the same time as the screenshot above was done. 1 second interval pings (the black line in the top is the result, a steady 2ms all the way through)
Edited by Irios - 21 hours 34 minutes ago at 12:00pm