![]() The results that you're seeing at the final destination in your traces are what you *really* want to focus on. If I trace directly to the router, it prioritizes those requests a bit differently (because the TTL on those requests isn't "0"), and I get clean results - same as you're seeing here. If I dial it down a bit (to a 2.5 or 5 second trace interval), it starts to respond a bit better (although, it will still show some packet loss here and there). I deal with this *exact* scenario with my connection at home: anytime I'm tracing to a target with a 1 second interval, my router shows 100% packet loss. We cover this concept in quite a bit more detail here: The important thing to keep in mind with this is: as long as the packet loss isn't carrying through to the hop directly after (or all the way through your route - to your final destination), then it's usually nothing you need to factor into your troubleshooting efforts. When they get these requests, they may down prioritize them (or not respond them at all) in an effort to make way for other network traffic. The first, and biggest thing to address here: some devices (in your case, your router) *really* aren't fond of timed out ICMP requests. Thanks for using PingPlotter! I'd be happy to help answer some of your questions here. Is there something I'm just woefully overlooking? Does anyone recognize this specific pattern?Įdit: This is what it looks like when I direct connect without the router: I've gotten to the point where I'm unplugging different appliances and playing with my fuse box to see if there's some dirty power or a bad ground introducing some noise but I can't seem to nail anything down and yeah, running out of ideas. What's especially odd is that in the first image you can see that there's packet loss in the first hop but if I run just that IP directly I get this without fail: ![]() I even seem to get good ping - and identical ping between "good" and "bad" signal. Using cmd, ping is always clean, no packet loss, yet I still see dicsonnects in software and PingPlotter seems to differentiate very specifically between when the line is clean and when something is amiss. I say artifact because I can't seem to see it with other tools. Any alternate configuration produces this artifact. The only thing that eliminates this completely and consistently is connecting a PC directly, via Ethernet, into the modem. The packet loss always shows up in one form or another. I've moved cables around, tried alternate NICs, multiple computers, tried wireless connections vs. ![]() I've also tried two routers, one brand new, with no difference. I've tested two separate modems with them on the phone, both looked great from their side but had identical issues from my side. I've spoken with my ISP and they can't see anything from their side, the line seems strong and clear. If the machine is idling with no network traffic it tends to disappear, but as soon as something gets launched, the cyclical loss appears and eventually causes disconnects in softwares that rely on stable connections - VoIP, games, etc. (The packet loss at 12:14am was a disconnected cable, I was swapping a new one in to test)īasically, intermittent, and seemingly cyclical 100% packet loss at the first hop, but clean after that. I've been dealing with some network issues for the past two weeks and I've run out of ideas as to what to troubleshoot.
0 Comments
Leave a Reply. |