JHuntVols Posted January 11 Share Posted January 11 I'm a bit frustrated here folks. I have a Nighthawk XR1000 router, and it has been working perfectly up until tonight when I resumed my gaming. For example, I was returning +0 on every bufferbloat test, and my online gaming felt amazing. Seriously chef's kiss. Lo and behold tonight it has gone to dog doo doo. I'm consistently getting +40 or more despite not changing any settings. I've even read a few threads and thought maybe disabling IPv6 on my AT&T fiber gateway would fix the issue, but it's made no difference. I had an epiphany when I was restarting my equipment though... the only change that has been made was me downloading a windows update. How could that affect my router performance though? Is that a possibility? Link to comment Share on other sites More sharing options...
JHuntVols Posted January 11 Author Share Posted January 11 Well, frustration is growing even more now. I did a system recovery to before the last update, but I'm still having the same issues. This is certainly not "bad reporting" on the part of WaveForm ... it's not coincidental to me that bad scores = bad game feel, while good scores = good game feel. It's a night and day difference to me, and it's again especially annoying when nothing else has changed yet it's simply not working now. Any help is appreciated! Link to comment Share on other sites More sharing options...
Administrators Netduma Fraser Posted January 11 Administrators Share Posted January 11 You are doing the test from that PC specifically so it is feasible it could change the results, you may need to re-fine tune QoS not to compensate, you should be able to get the same results back. Also Traffic Prioritization can interfere with the results so best to disable it while testing Link to comment Share on other sites More sharing options...
JHuntVols Posted January 11 Author Share Posted January 11 4 hours ago, Netduma Fraser said: You are doing the test from that PC specifically so it is feasible it could change the results, you may need to re-fine tune QoS not to compensate, you should be able to get the same results back. Also Traffic Prioritization can interfere with the results so best to disable it while testing I'm starting to wonder if this might be an AT&T problem again. I'm simply not getting full speed (1GB) even with QoS disabled. When it is enabled... the speed is noticeably inconsistent. I did retest with other sliders (15%, 30%, 40%), but I got no improvement regardless of which way I went on the slider. I also did disable traffic prioritization as you suggested. Is it possible AT&T is updating their gateway and changing settings without me noticing? Link to comment Share on other sites More sharing options...
Administrators Netduma Fraser Posted January 11 Administrators Share Posted January 11 59 minutes ago, JHuntVols said: I'm starting to wonder if this might be an AT&T problem again. I'm simply not getting full speed (1GB) even with QoS disabled. When it is enabled... the speed is noticeably inconsistent. I did retest with other sliders (15%, 30%, 40%), but I got no improvement regardless of which way I went on the slider. I also did disable traffic prioritization as you suggested. Is it possible AT&T is updating their gateway and changing settings without me noticing? They could be, it's quite possible. Easiest way to tell if it's an issue on their side is test direct to the ISP modem/router, do you get expected speeds? Link to comment Share on other sites More sharing options...
JHuntVols Posted January 11 Author Share Posted January 11 8 minutes ago, Netduma Fraser said: They could be, it's quite possible. Easiest way to tell if it's an issue on their side is test direct to the ISP modem/router, do you get expected speeds? I don't know who to believe honestly... speedtest is telling me no, but the gateway is telling me yes. Link to comment Share on other sites More sharing options...
Administrators Netduma Fraser Posted January 11 Administrators Share Posted January 11 The gateway will test directly to the ISP so it won't hit the open internet which is why it tests higher. Speedtest.net is more accurate of what speeds devices will see, try other speedtest sites just to make sure Link to comment Share on other sites More sharing options...
JHuntVols Posted January 11 Author Share Posted January 11 They're all around the same now... mid to high 700s on download, then I'll see high 700s to high 800s on upload. Link to comment Share on other sites More sharing options...
Administrators Netduma Fraser Posted January 11 Administrators Share Posted January 11 Either an issue on the ISP side then or it's just the max you can achieve at the time, the speed may vary throughout the day depending on the time you test it Link to comment Share on other sites More sharing options...
JHuntVols Posted January 12 Author Share Posted January 12 4 hours ago, Netduma Fraser said: Either an issue on the ISP side then or it's just the max you can achieve at the time, the speed may vary throughout the day depending on the time you test it Are you saying that's what causing my bufferbloat? Link to comment Share on other sites More sharing options...
JHuntVols Posted January 12 Author Share Posted January 12 Here's my pingplotter results BTW. If I'm reading this correctly, then the problem is packet loss on my side, eh? Link to comment Share on other sites More sharing options...
Administrators Netduma Fraser Posted January 12 Administrators Share Posted January 12 7 hours ago, JHuntVols said: Are you saying that's what causing my bufferbloat? 4 hours ago, JHuntVols said: Here's my pingplotter results BTW. If I'm reading this correctly, then the problem is packet loss on my side, eh? Not necessarily but if your telling the router it's getting more speed than it actually is then its using those values for QoS. So for example, seems your max speeds are ~700ish, you tell the router you have 1000Mbps and you drop the percentages to 85% it's doing nothing because it's not actually got to your speeds yet. That would throw out any changes you make to get your bufferbloat down, if you correct that you should get better results. No, the bottom graph is the part to focus on and that doesn't show any packet loss. Link to comment Share on other sites More sharing options...
JHuntVols Posted January 12 Author Share Posted January 12 8 hours ago, Netduma Fraser said: Not necessarily but if your telling the router it's getting more speed than it actually is then its using those values for QoS. So for example, seems your max speeds are ~700ish, you tell the router you have 1000Mbps and you drop the percentages to 85% it's doing nothing because it's not actually got to your speeds yet. That would throw out any changes you make to get your bufferbloat down, if you correct that you should get better results. No, the bottom graph is the part to focus on and that doesn't show any packet loss. Well, I'm just going by what their website says... I get your point but wouldn't dropping it down to 15% account for any margin of error in the connection speed? I'd presume so, and I still have the packet loss problem too. Link to comment Share on other sites More sharing options...
Administrators Netduma Fraser Posted January 12 Administrators Share Posted January 12 It would definitely be applying at that point but it would be better to just tell the router what speeds you're getting so QoS can be more effective and you also can't go below 50% for Congestion Control currently. Here you go: https://www.pingman.com/kb/article/initial-hops-100-packet-loss-6.html Link to comment Share on other sites More sharing options...
JHuntVols Posted January 13 Author Share Posted January 13 Fraser, I have a BGW320-500 gateway from AT&T. I have a 1GB Fiber plan. I disabled the DHCP server on my gateway thinking it may be causing IP Address conflicts with the understanding that function was now only to be done by my router. Unfortunately, I lost the ability to connect to either the router or gateway. I had to factory reset both to be able to access the gateway again. I obviously touched nothing on the gateway side, then I let the router set itself up the wizard. It all worked again, so I decided to attempt to redo the settings as it was before disabling the DHCP Server on the gateway. This included bandwith allocation on both the application and device side, setting up QoS with speeds confirmed to by SpeedTest to be in the low 900s on both the upload and download for the relatively simple things. I then went into the advanced settings to manually select "Use this MAC Address" for the Router MAC Address (confirmed to be the router BTW). I changed nothing in the WAN or LAN setup, and I also noticed that "Use Router as DHCP Server" was checked as expected. I went to the gateway and set the device to be in IP Passthrough with the mode being DHCS Fixed and tied to the router's MAC Address. I also disabled all WiFi on the gateway. The last change I made was in the router to the "Internet Connection Type" to be Pass Through before applying it and restarting both the router and gateway. Alas, the screenshot was my result from Bufferbloat. QoS was not enabled for the test to be clear. Any ideas? Link to comment Share on other sites More sharing options...
Administrators Netduma Fraser Posted January 13 Administrators Share Posted January 13 The function of DHCP is to hand out IP addresses to connected devices so if you disable this without having set a static/reserved IP the gateway can't give an IP to a device and therefore you won't get internet, that's why that happened. The unloaded ping is good so there isn't an inherent problem, the active ping is bad so it does indicate that you need to enable QoS and fine tune it to get those values down. Link to comment Share on other sites More sharing options...
JHuntVols Posted April 19 Author Share Posted April 19 In case anyone is having a similar issue, then I'm happy to say I found a solution. I got rid of the XR1000 Nighthawk and bought a Ubiquiti gateway / router. I've since had 0 issues with again the only thing changed being the gateway / router. The screenshot of the bufferbloat test below is honestly the worst I've seen it; smart queues is enabled during this test. Thanks. Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now