After moving house I have been struggling with my DSL and blamed many issues on my provider (Vodafone)

But I finally made progress and realised some of the issues I had were not issues...

As it stands right now, everything appears to be stable.. And I can do a 'dos' ping to 8.8.8.8 for hours on end with zero packet loss.

But I have been using PingPlotter for my tests and that tells a different story.. With PP I am getting regular bursts of packet loss.. Further more, when its running my "dos" ping starts to fail too.

It looks like I get packet loss across the network but not to my first and 2 hop.. So what is this telling me? I guess if there is a problem its not at my end..

Does this show a problem? Or is this a problem with PP? And if its correct, why does PP show this, but a DOS ping not?

I would really appreciate someones help with this as I am not sure if I have a problem or not.

Jon

I may have worked it out... I was running a 1 second ping.. I just did the same on another PC and its exactly the same.

So I reduced it down to 5 seconds and now its a lot more stable.

Could this be some kind of anti-Denial Of Service thing where the network it doesn't allow back to pings?

Hi Jweaver,

Thanks for writing into the forum!

Without seeing your data it's hard to be 100% certain but from what you describe it sounds like you hit the nail on the head!

Some routers are configured to down-prioritize ICMP echo requests where the TTL = 0 when it reaches the router and instead prioritizes "Real" traffic. They'll generally pass the packet along just fine but will drop it and won't respond if there is too much traffic. See this article for more details.

Some devices also have a Denial of Service mechanism that can get triggered if they determine that they're getting rapidly bombarded by ICMP packets from the same IP address. This sounds likely in your case since it settled down after changing the Interval to 5 seconds.

The most important thing is the final hop, if there's packet loss or high latency at the final hop then you need to trace it back through the route to see where it originates. If the final hop is good you can disregard all of the other packet loss and latency since it's likely due to the previously mentioned behavior.

Let me know if this leaves you with any questions!

Thanks,

-Poe

Many thanks Poe...

I am pretty much sure there isn't an issue... After reducing the interval to 5 seconds its now been pinging 8.8.8.8 with zero packet loss for the past 6 hours.

I did bring up a graph by accident of an intermediate hope which has the odd dropped packet.. But its so infrequent that the loss still shows 0%

So I am pretty sure that I will simply sending too many pings.. And now I have slowed it down to the same rate as a "dos" ping it looks good.

Thanks for taking the time to reply.

Jon