Pingman Pro V5.11.5.5106
ISP Cogeco.ca cable 120Mbps.

The first hop alternates between 182.168.0.2 (named puma7-atom.cogeco.local), which I presume is my modem's address, and 10.75.224.1 (unnamed).

The second hop is also 10.75.224.1. As you will see, it looks like a ping is not shown from 10.75.224.1 on Hop 2 when one is received on Hop1.

Can anyone explain why this is happening?

Hi Frank,

Thanks for posting in the forum!

Thanks also for providing a good explanation of what you're seeing and a screenshot, very helpful. I'm happy to give you my two cents on what I'm seeing based on your data. That said this screenshot only represents about a minute of time and it's important to trace for a long period of time (we suggest 48 hrs) before being able to confidently establish what and where the issues are.

In your screenshot, I see that the first four hops are all internal private addresses (10.0.0.0 – 10.255.255.255 and 192.168.0.0 – 192.168.255.255) and the first public address (209.85.173.40) is Google. This is interesting but it may just be how your ISP (Cogeco.ca) has set things up. The latency to 8.8.8.8 (24.1 ms) is pretty good so I wouldn't worry about that too much. What I would suggest looking into is the packet loss that starts at hop #1 and goes all the way through to the destination.

This means that the issue is likely originating at your router (192.168.0.1) and any packet loss, even 4.2%, can be problematic. It sounds like your router is separate from your cable modem so one thing you can try is plugging your computer directly into the modem (with an ethernet cable) and see if the packet loss is still present. If it's gone when connected directly to your modem than that means it's being introduced by your router. If this is a consistent issue you may need to factory reset (Google the manufacturer and model number with the words "factory reset") your router. If the issue persists after a reset you may need to actually replace your router. If you lease the router from your cable company you can usually just ask for a replacement.

If this issue still exists when just plugged into the modem than that suggests that the packet loss is being introduced by the modem rather than the router. Again, if you lease the equipment from your cable company you can usually just ask for a replacement.

If you want to trace for 48hrs and send us the results we would be happy to take a look at that as well, just create a share page (File -> Share -> Create Share Page) and send the URL for the page to support@pingman.com.

Let us know if you have any further questions, we're happy to help!

thanks,

-Poe

I just gave a short time period (of a scan of over 36hours) so that the first line would clearly show the alteration of the local 192.186.x.x. IP pings and the first 10.x.x.x pings. Perhaps what I didn’t clearly state is what’s going on in line 1.

When a ping is recorded in line 1, it shows the ping as coming pack from 192.168.x.x (puna7-etc.) at the left end of the line. On the next ping, ithe legend at the left end of line 1 shows that it is dealing with pings from 10.x.x.x, I.e. the same IP as line 2. When this happens, a ping is recorded in line 2, but line 1 is just greyed outfit that cycle. This alternation continues but there are the odd cases, one is shown on the chart I provided, where doesn’t happen.

Put another way, on one cycle, a ping is recorded from 192.168.x.x only on line 1 and on the next cycle, a ping is recorded from 10.x.x.x only. (Note that when a ping is recorded on line 1 from 192.168.x.x, no ping is recorded on line 2.)

As for packet loss, I have pinged 192.168.x.x for a time with no packet loss being recorded. During the cycle where packet loss was recorded was recorded, line 1 was actually looking at 10.x.x.x.

I was not really asking what’s going on with my Internet service. Rather, I was wondering why PingPlotter was alternating between two IPs.

Initially, I got PingPlotter because I was having trouble with my Bell DSL service. It turned out to be a noisy line and they couldn’t find a good one. So I moved to Cogeco. The service is generally very good, but they acknowledge that they have an unlocated intermittent fault somewhere on our street. Over a period of about 15 minutes, the packet loss would go from 0% to 100% on hop 2. After waiting a while, resetting the modem generally restored service. I notice that the the IP address for hop 2 changed a couple of days ago, so perhaps the problem is now fixed.

Here's a video which hopefully will better explain what is going on.

http://thehowards.ca/images/pingplotter2.mov

You will see that in the list if IP addresses, hop 1 alternates between 192.168.0.1 (presumably the modem) and 10.75.224.1 (the same IP as hop 2).

I can't replicate my earlier experience where Hop 2's chart would not show a ping when hop 1 is pinging the modem. Well, as I write, that has come back after a fashion, pinging 8.8.8.8. But now I get three pings of both hops 1 and 2. On the next ping hop1 is greyed out and hop 2's ping time goes from an average of about 10ms to somewhere close to 200ms.

This behaviour happens for about 1m30s then stops for about 2m on a very predictable cycle.

What could be causing this in the network, or is it an artifact within PingPlotter?

Hi Frank,

Thanks for the video!

You know what they say about a picture being worth a thousand words, well a video has to equal at least 10 emails/posts.

This is a very strange behavior and a very strange route as well. Do you notice that when the first-hop changes IP addresses in your video it also adds in another hop at hop #9 and then it disappears when it switches back? It also shows gray periods in the timeline graph for hop #1 which basically means that PingPlotter didn't expect a result for some reason.

Also when I do a WHOIS lookup on the first external IP address at hop #4 (64.34.137.61) it's actually registered to the Toronto Star Newspaper. Why are you being routed through an IP that belongs to a Newspaper?

We have seen similar IP switching behavior once or twice before but because they were on customer networks we were never able to figure out exactly what causes this behavior. I can pretty confidently say that this is not a PingPlotter aberration as I believe that the ISPs were able to fix the customer's issues.

I'd recommend getting ahold of Cogeco and sending them this video and asking them what's going on. If they give you a hard time or try to explain it away I'd just ask for their manager until you get someone who understands routing enough to know that this is weird and it's likely in their network.

Good luck to you and if you figure it out we'd certainly love to know what causes this kind of behavior.

If you have any other questions we're happy to help out however we can!

Thanks,

-Poe

Thanks, Poe. I’ll see what Cogeco have to say about it.

Another add thing in this area is that sometimes PingPlotter doesn’t list modem in its routing a all and the first two hops are the same 10.x.x.x address. The two charts, however, show different ping results.

Incidentally, I find that I get a very similar behaviour using PingPlotter for iOS V1.3.1.

I have wondered that same thing regarding use of the Toronto Star’s node. But then, Cogeco is a Cable Company and may have less than traditional partnerships.

Hi Frank,

Thanks for the update!

Let us know what Cogeco says - we would certainly be interested to learn a little more about what might cause the results you're seeing.

Let us know if you have any other questions, we're happy to help!

Thanks,

-Poe