Pingtest.net

[mechmike0034]

Pingtest.net (http://pingtest.net/) is a free broadband quality analyzer designed to complement Speedtest.net. This troubleshooting tool grades your connection on how well it is likely to handle things like voice services (VoIP), streaming music or video, telecommuting or online gaming.

Another one for the bookmarks...

I hope that at some point they do a mini version (like Speedtest.net) that you can host free on your webpage or blog.

[ray]

I found the test on pingtest.net took longer to actually run than when I run it with Speedtest.net. I don't know if it's due to the website design itself (slower loading pages) or perhaps my connection to the pingtest servers are actually slower than my connections to the speedtest servers. In any case, nice to have another option to get a gauge on performance.

[Cheesypuff]

http://www.pingtest.net/result/4186946.png (http://www.pingtest.net)

this is mine =) it sucketh

[Airencracken]

I got a B*. (I don't have a JRE on my computer)

[mechmike0034]

http://www.pingtest.net/result/4236701.png

That's from the notebook while listening to the most excellent W3 Blues Radio (http://translate.google.com/translate?hl=en&sl=fr&u=http://w3bluesradio.com/&ei=sF4LS7LBHd7VlQeEpu3lDA&sa=X&oi=translate&ct=result&resnum=1&ved=0CBAQ7gEwAA&prev=/search%3Fq%3Dw3%2Bblues%2Bradio%26hl%3Den%26rlz%3D1C1GGLS_enUS314US314) streamed to the desktop...

[gwilks98]

http://www.pingtest.net/result/4186946.png (http://www.pingtest.net)

this is mine =) it sucketh

my guess it's because of packet loss plus jitter. 2% packet loss bites, but it's still usable. Anything above 5% and you'll feel like you're on dialup.

[gwilks98]

Good find Mike.

As a networker, I launched Wireshark to take a look at what this guy's doing during the test.
During the "ping test" he's not pinging at all, but sending UDP to the target server on 8080, all one directional traffic from me outbound. (Ping is an ICMP request followed by an ICMP response on a different port.) I wasn't expecting his result to be accurate since the traffic was all in one direction, (and ping is two directions) but he was spot on.
During the "jitter test" he's sending what appears to be near-blank sequenced non-HTTP data on 8080 to the target, receiving a response, then repeating ad-nauseum. Each direction is sequence per TCP and I'm betting he's measuring how often packets arrive out of sequence.

I'm curious how a test like this would work inside my work, where we have IDS/IPS sensors, internet compression devices and a slow proxy server getting in the way.

It'd be neat if they would allow you to adjust the time the test can run. 5-10 seconds won't paint a very accurate picture on a flakey link.


Something else to mention: this test doesn't explain the whole picture. If you ISP supports it, you can flag your voice/multimedia/streaming traffic with QoS flags which can get that traffic preferential treatment on the wire. This ping-test will not be able to account for this, but I don't think it's designed to. :)

[Airencracken]

Good find Mike.

As a networker, I launched Wireshark to take a look at what this guy's doing during the test.
During the "ping test" he's not pinging at all, but sending UDP to the target server on 8080, all one directional traffic from me outbound. (Ping is an ICMP request followed by an ICMP response on a different port.) I wasn't expecting his result to be accurate since the traffic was all in one direction, (and ping is two directions) but he was spot on.
During the "jitter test" he's sending what appears to be near-blank sequenced non-HTTP data on 8080 to the target, receiving a response, then repeating ad-nauseum. Each direction is sequence per TCP and I'm betting he's measuring how often packets arrive out of sequence.

I'm curious how a test like this would work inside my work, where we have IDS/IPS sensors, internet compression devices and a slow proxy server getting in the way.

It'd be neat if they would allow you to adjust the time the test can run. 5-10 seconds won't paint a very accurate picture on a flakey link.


Something else to mention: this test doesn't explain the whole picture. If you ISP supports it, you can flag your voice/multimedia/streaming traffic with QoS flags which can get that traffic preferential treatment on the wire. This ping-test will not be able to account for this, but I don't think it's designed to. :)

Makes me wonder if snort interfered with my test.

[ramazank2]

Wow I got a nice connection at home!!!
http://www.speedtest.net/result/632352674.png

My first scan came out horrible....
http://www.pingtest.net/result/4268272.png

A retest came out great! Can I really trust a site that varies so much from test to test?
http://www.pingtest.net/result/4268381.png

[Markel]

No complaints here

http://www.pingtest.net/result/4273173.png (http://www.pingtest.net)

[gwilks98]

A retest came out great! Can I really trust a site that varies so much from test to test?

You sure it's the site and not your connection?

[mechmike0034]

You sure it's the site and not your connection?

Exactly - run multiple tests if you question the first test's results. Try a couple of different servers as well. You're looking for long-term trends, not short-term spikes.

[gwilks98]

One more thing...it could be your router. Home (wireless) routers are notorius for not being able to handle consistent streams.

[ramazank2]

Nobody talks **** about my router... It is bad ass!!! (linksys 610N)

Just playin... I dont know what happened in the first scan.