Speakeasy → Re: Pingplotter

Hi everybody:

I opened the program several times and I can do multi pings that way. It's interesting to see when I have a dropout on just one IP or several. To see if the times coincide.

Hi Tim: I'll put in the IP address you gave and see how it does.

I was pinging the gateway IP of SE (really Covad), which is 66.92.238.1. I guess that gateway is at my local CO. I assume the two dns servers are also at my local CO?

It's getting closer to the time that I usually get dsl drop outs (about 7 PM to 9 PM). I did see longer drop outs as it got around 6 PM. Most of my drop outs earlier in the day were only about 5 seconds.

Hi cmaenginsb: I can't wait til I can understand the more complicated tools like you have...lol. I'm just lucky to get Pingplotter working...lol. It actually turns on an audio alarm when dsl goes off. I can hear it downstairs. So I don't have to keep my eyes glued to the computer to know when an alarm occures.

Hi ILC: I'll take a look at MultiPing. Thanks.

Hi Airplace777 -

Speakeasy actually runs its own private network, and only uses Covad for last mile connectivity from the CO to your customer premises.

All that's housed at your CO is our last mile provider's DSLAM equipment, and trunking to an upstream BPX. The BPX aggregates all ATM traffic, to hand off to Speakeasy on our backhaul.

More information is available at »www.speakeasy.net/network if you're interested in inter-POP latency, and white papers published by Speakeasy.

The IP address, 66.92.238.1 is the Juniper ERX housed at Speakeasy's NYC POP. This definitely wouldn't be the right IP to use for testing purposes, as the Juniper's will prioritize ICMP as the lowest possible, as actual data traffic deserves a higher QoS.

It's best to use your DNS server, 216.254.95.2 for testing purposes, as the Juniper will treat packets destined for end points beyond it with priority.

If you are in fact experiencing intermittent drop outs, and it's related to the last mile connectivity, training starts would also increment on an ADSL circuit, and may be the best statistic for determining if an issue is occuring with the physical loop, or if the problem is further down the line.

Thanks,
Adam

Thank you so much for your post.Hi adamA:

I used the 66.92.238.1 IP address since the SE support guy on the phone last night told me it was at my local CO. But after looking at several drop outs today, I wondered why I didn't see a drop out on that IP address (that I thought was local) coincide with drop outs on some other IP addresses. So what you say makes sense, that my SE gateway is not really at my local CO.

Is there any equipment at my local CO that I can ping just for my local loop?What do you mean when you say "training starts".

How do I see these "training starts"? Isn't there any kind of IP I could ping at my local CO to help me measure just the local loop?

Thanks

There isn't any layer 3 equipment at the CO itself you could use for ping testing alone. All of it is ATM level. As far as your IP experience goes, your first hop (our gateway) traverses from your machine, is converted to an ATM cell by the DSL CPE, and remains an ATM cell all the way to the backhaul handoff at our Juniper.

The Juniper (our gateway) then converts the cell back to IP packets, and routes it to your destination.

Training starts are incremented on the DSLAM anytime the connection loses sync and the DSL CPE attempts to retrain.

Speakeasy has visibility into the number of training starts on your circuit-- this information is easily accessible, and can be provided to you by us on request, either by phone or an online ticket.

Some users on the forums here have posted instructions on how to log in to DSL modems to see training starts incrementing, are you using a Broadxent 8012?

Thanks,