Hi there

I'd like to start off by saying that PingPlotter is a great program, without it there would be no hope with my ISP. Awesome job you guys have done on it.

Anyways, I've collected quite a bit of data running multiple PingPlotter tests to different servers. 1 of my favourite game servers (where the lag first reared its ugly head) both my primary and secondary DNS servers and a BF2 server that runs on a network that my ISP peers with. I would like to post my PingPlotter data on here for u to have a look at but i couldnt work out how to properly configure my personal web page lol. I tried for over an hour but alas. Maybe I could email them to be hosted or just for someone to have a look at them?

I've already contacted my ISP after I had collected a much smaller amount of data than what i've got now and I got an email back from them today saying that there was no apparent faults (which is totally wrong, there *is* something wrong there). Is it normal for them to disconnect your adsl for them to run any kind of test? As i said before i've been running PingPlotter for a while and it shows that there was an hour and a half worth of packet loss, I checked the graph that was my adsl modem/router for that time period and there was no packetloss there so it wasnt my router locking up as it had done the night before.

Also about 3 hours or so after that insane amount of packetloss (disconnection) my route changes, but only with 2 extra hops. Both extra hops are both my ISP's servers so im guessing that this is no coincidence?

Neways i'll wait for someone to be able to see these screenshots before babbling more stuff that prolly doesnt mean anything without a screenshot

Hope someone can help
Haynsey

If you want to attach an image, the message board here will allow you to attach a .gif, .png or .jpg. If you want to send us a .pp2 and we'll post some images, send an appropriate .pp2 file to support@pingplotter.com and we'll review it and post back here.

.pp2 file sent to that email address. By the way, that pp2 is of a gaming server in the USA and as you can probably tell from that file that I am in Australia. It doesnt matter so much about the round trip time as much as what amount of latency hops 4 between 5 before the route changed, and hops 6 and 7 after the route change.

It looks like an overloaded link as it only happens during the times of day/night you would expect network usage to skyrocket. Its fine otherwise, except the only time i mean by otherwise is between 3 at night and 11 in the morning, any other time means an extra 100ms tacked onto anything you do as all my information has to go throught that saturated link of theres. May not sound like too much when browsing the net, but with added packetloss and the lust for online gaming it is extremely detrimental.

Im only guessing really, im pretty new to all this stuff. What do you think about that .pp2 file?

Haynsey

I did some analysis of your data and have attached a picture that captures a few separate problems.

First, on 11/9, 8:37pm, your router locked up. You made note of that (nice work!) and it's pretty obvious because hop 1 stopped responding. Since hop 1 is entirely on your network, packet loss at hop 1 is localized to hardware under your direct control. Note that there was some packet loss in the period before the router locked up - I'm not sure why it locked up, but for the period before lockup, it decided it showed quite a bit of packet loss that did not affect downstream hops. This means that your router decided to stop responding to ICMP TTL=0 packets, but to pass everything else downstream. It's an interesting pattern, but isn't the problem you were asking about, so I won't spend any more time on this.

There is a pretty bad latency curve, along with some packet loss, from the period at the beginning of the graph all the way through about 3:00am on the 10th. I focused the upper graph on this period because it's relatively interesting. Hop 4 is nice and solid, while hop 5 has packet loss and latency - abnormally so.

One possible reason for this is if there's not enough bandwidth to service all the customers. Hop 5 is probably a peering link, and the bandwidth between hop 4 and hop 5 is probably something that your ISP has to pay for. It looks like there's not enough.

The second and third packet loss periods you noted all start with the first hop inside your ISP (a private address, which is a bit odd!). They may have had equipment problems and they were working on it (although the outage at 6:00 pm is pretty long). I can't really speculate on what caused that - you'll need to have your ISP explain that for you.

At roughly 12:45 pm, you did get two more hops inserted between hops 4 and 5. Maybe your ISP decided to insert some monitoring devices, or maybe they reconfigured some of their hardware to decrement the TTL on packets. Again, it's hard to speculate what happened. There really isn't anything odd happening before those two hops were inserted.

I would definitely be looking at your ISP here for a solution. The long packet loss periods are almost certainly their doing, as is the latency problem.

Thanks for the quick and comprehensive reply

I shall be looking into this further through my ISP, I didnt think such a bad latency spike during digital peak hour was normal but as I said before all of this is relatively new to me.

At the moment my pings are looking pretty good, either whatever they have done has improved it or the network aint in much use yet, I think i'll wait a day before making my case to the ISP, a succesful one hopefully hehe

Thanks for all ya help mate, I'll b recommending PingPlotter to anyone that has a network problem. Once again, great program.

Haynsey