XR1000 ping spikes, poor connection and bufferbloat

I've attached a screenshot of my Waveform bufferbloat test and a few others along the way considering I can only add one at a time.

I've now defaulted everything and set up my router so that I can try to find the sweet-spot to hopefully lower bufferbloat, that way I plan to allocate that speed to my gaming PC, and the rest to my Streaming PC which I hope isn't an awful logical approach, regardless I'm still the only internet user in this house.

Here's my Geo-Filter for the games I play.

The benchmark in the router.

These are my congestion Control settings with sharing excess disabled, I've set it up like this so that I can dedicate everything to finding a good setting to reduce bufferbloat, I've set my download speeds as the highest speeds a speedtest can do which is usually around 490 download and 49 upload, rather than my 500/50.

So my settings are good so far? Upload/Download Share Excess both disabled to make proper use of congestion control, seems like I've gotten some stability so far, I'll play with the slider again and see if it works this time around.

Thanks a lot for the help so far by the way, I'll report back once I'm finished.

I've now been through every single increment of 5% on both download and upload over 2 hours, one actually gave me a good 95'th percentile, the in-built router benchmark shows almost the same results everytime with mild variations but the waveform bufferbloat just doesn't really budge unless I do a benchmark at the right time.

The one good result I had could not be recreated when repeating the test with the same settings and I'm not sure why there's such massive variations in ping.

I've linked my best test result which was at 35mbps/5mbps from the in-built benchmark in attachments.

I've just done a test at maximum speeds with an without QoS enabled as well as goodput disabled, it looks like QoS is not working as intended.

Above is with QoS enabled, below is QoS disabled.

I'll test it with pingplotter instead, but pingplotter doesn't show the same disturbing spikes which I think cause my ingame de-sync, also I still have a modem which is on a list of bad modems, I think it plays a huge role in this case. It's an arris tm3402.

Thats exactly the plan so far, both me and my mate have Puma6/7 at the moment and we're going to try and see if there's a way to get rid of that first, I'm fairly confident that it's exactly that being an issue at the moment. We'll see. I'll report back when I've got something new to test.

So far I haven't been able to get any desireable results whatsoever.

Hey again, I've just visited my ISP with a huge complaint against the selection of modems and routers and it turns out that my ISP only has Puma 6 and Puma 7 chipset modems and routers, I've gotta go and check if I can somehow get rid of it, but prior to that do you have any way of helping me figure that out? I'm under the impression that my TM3402 might bottleneck my network a lot, but I'd like to confirm and prove that to perhaps usher my ISP towards helping me fix that, we're getting some help on our RFoG connection to confirm that it works.

So far I haven't been able to enjoy my XR1000 and I think it's possible if I get to solve this modem problem I have, but I have no idea how to go about it since I don't know which servers to ping and what kind of result to expect.

I don't hope this is too much to ask for, but I'd be extremely thankful if there's something you could assist with, so far I'm still extremely happy with the support so far!

Hope you've had a wonderful christmas regardless!

Thank you.

Well, it's really hard to tell, I've been down there to ask for ANY possibilities of getting rid of this specific chipset. We'll see how that goes.

As a closing note before I get an update again, this router should most definitely have an effect on bufferbloat one way or another, right?

I've been through every position with the congestion slider through 100% to 0%, neither of them provided anything below 10-20, sometimes 7% would give +10 download latency and another result with the same setting would be a +20.

With that being said it's safe to assume that something is wrong that the XR1000 can't control, right?