Topic Options
#2536 - 08/04/14 09:33 PM What would cause this pattern of packetloss?
Dangime Offline


Registered: 08/04/14
Posts: 1
I live in a rural area where a line of sight WISP is the only real option for internet service. I had a serious problem with ping spikes, so I started running ping plotter.

What I saw was a low amount of packetloss from the closer hops which I kind of expected. It's never perfect due to the nature of the system. Also it gets really bad during prime time, but I figure this is just their system being overloaded during those times.

The thing that really stood out to me is the constant 15% or so packetloss listed from the middle IP ranges controled by the bandwidth provider of the ISP. These stand out as being a constant problem 24/7. I tried to contact my ISP about the problem, but they were dismissive of the idea that the middle man might be the problem and only offered to look at my side of the connection. Honestly, I'm afraid to even let them touch my side because I work from home and a jinky connection is better than no connection, but is there anything you could tell me about this packet loss with the middleman that might reveal more? Maybe I can convince my ISP it's an issue.

Oh...I'd like to post the data, but it looks like I've collected too much to attach it. Here's a screen shot. Thanks for any help you can provide.

[img]http://imgur.com/ip6kMYM[/img]


Edited by Dangime (08/04/14 09:34 PM)

Top
#2538 - 08/05/14 11:37 AM Re: What would cause this pattern of packetloss? [Re: Dangime]
Gary Offline
PingPlotter Staff


Registered: 10/30/13
Posts: 185
Hi Dangime,

Thanks for writing in!

You've definitely got some interesting results here. From what I can see in the screenshot you attached to your post, the packet loss you're experiencing seems to be coming from somewhere pretty close to home (but it sounds like you already knew that).

The middle hops (10, 11, and 12) that are showing higher packet loss aren't necessarily contributing to your issue here - as the pattern they present doesn't appear to be continuing through the rest of your route. Some devices/routers simply don't prioritize timed out ICMP request very high (or don't respond to them at all). You can see a more detail description of this topic here:

http://www.nessoft.com/kb/5

It's tough to diagnose exactly what may be happening here based off of a single screenshot... but from the looks of it, the culprit is most likely the device at hop #1 (or something between the computer running the trace and that device). This is *probably* the best place to start your troubleshooting efforts.

We've got a newly released guide that goes over some of the best practices for troubleshooting scenarios like this (and offers advice on how to build a case to present to your provider if necessary as well) which may be helpful. Have a look here:

http://www.pingplotter.com/netnirvana/

Hopefully this helps get things headed in the right direction! If you have any questions, or find yourself needing any assistance - please don't hesitate to let us know.

Best wishes,

-Gary

Top

Search

Who's Online
0 registered (), 18 Guests and 1 Spider online.
Key: Admin, Global Mod, Mod