Jump to content

XR500 Performance


Recommended Posts

Just running some tests to give people general information and try to figure out my best configuration.

 

Cable Internet (personal NetGear CM600 cable modem)

300Mbps Download (usually see 350Mbps)

20Mbps Upload (usually see 25Mbps)

Wireless disabled due to having an AC WAP. This may change performance if the WiFi uses the main CPU, I have not looked in to the full specs of the Netgear. I know ASUS has a separate CPU setup for the WiFi so this needs to be taken in to account.

 

All QoS settings Disabled (Still received an A for BufferBloat and A+ for Quality) Curious as to why not reaching 350Mbps from DSLReports when getting it from Speedtest with same settings. These tests would be much more accurate if the tests had equal results.

29450019.png

 

7037952341.png

 

With Anti-BufferBloat enabled

 

JziVHm7.png

 

 

29450727.png

 

7037983346.png

 

Changed Anti-BB to "Always"

 

29450865.png

 

7037994233.png

 

Adjust sliders to 90% (still on "Always")

 

29450972.png

 

7038005598.png

 

 

 

The XR500 handles my connection pretty well even without QoS enabled. Granted, for some reason, DSLReports was not getting close to what Speedtest gave me in terms of download so that could make a difference because Speedtest doesn't show me BufferBloat results and DSLReports isn't reaching a true indicator of what my connection is.

The best thing, so far, is the GeoFilter. It has shown some serious issues, to me, with America's Army matchmaking, server selection and allowable connections to their servers. I am sorry, have players around the Midwest, connect to an East server with a player from Colombia and Saudi Arabia is going to cause issues! I still have not figured out why developers don't understand basic physics or are they just so arrogant that their algorithms are perfect? But this tool is the easiest way to troubleshoot for users.

 

For now, I will be keeping my sliders at 100% (with Speedtest,net speeds entered) , "When High Priority" selected because the NetDuma seems to handle the load just fine. I would think, any cable, internet under 300x 20 could do the same without issue. I never saw a BufferBloat grade less than A, even with no QoS enabled, that is huge for gamers! Me not using the XR500 for WiFi needs to be considered with these findings. Using WiFi may take resources of  the CPU that are giving me my results.

 

Definitely VERY happy with my purchase.

Link to comment
Share on other sites

  • Administrators

Hi Shane, DSL uses port 80 to run their tests. Currently that is the only port not hardware accelerated by the router due to the blocking features in settings. This is why the speeds don't match speedtest.net. We'll be implementing something in the next firmware to allow hardware acceleration on port 80. I think you would gain a better experience by lowering the sliders more below 100% such as at 70% with 'Always' selected.

Link to comment
Share on other sites

Well I just ran multiple test's, and all my results were as expected. I tried with having QoS disabled, and with it enabled. I wonder under Bandwidth Allocation, do you have the option of "Share Excess" enabled? Because if not, whatever limit's you have set, and the percentage you're giving that device under Bandwidth Allocation, it will only be able to use that much, with "Share Excess" being disabled.

 

I don't know what kind of internet connection you have though, I have a cable provider for my internet.

Link to comment
Share on other sites

Hi Shane, DSL uses port 80 to run their tests. Currently that is the only port not hardware accelerated by the router due to the blocking features in settings.

 

I see this option on Dslreports.com, however mine isn't enabled. I was still able to hit my speed's, and over. With QoS enabled, and anti BB set at 99% on both upload, and download. It limited my speed's to what I had set overall. So for me at least, it's working properly without needing to use port 80 only.

Link to comment
Share on other sites

I have "Share Access" checked.

 

Thanks Fraser for that answer.

 

**Edit**

 

On the DSL page you can change some settings and changed it to test using HTTPS and I am reaching the proper speeds.

 

"High Priority" and sliders at 100%

 

29453509.png

 

 

"Always"

 

29453570.png

 

"High Priority" with sliders at 95%  - I have a feeling this will be my "sweet spot"

 

29453642.png

Link to comment
Share on other sites

  • Administrators

I have "Share Access" checked.

 

Thanks Fraser for that answer.

 

**Edit**

 

On the DSL page you can change some settings and changed it to test using HTTPS and I am reaching the proper speeds.

 

Nice one, thanks for letting us know. Didn't know you could change it to HTTPS!

Link to comment
Share on other sites

Does DumaOS have the preemptive and reactive options? Interesting to see some quality grades below A+.

 

The reason I ask is because if I use reactive or device prioritisation on the R1, DSLReports reports packet loss as shown by a quality grade other than A+. If I use both reactive and device prioritisation with share excess disabled, there's a larger amount of packet loss. If I do that and then drop my sliders right down (like <30% bandwidth), I can end up with as low as an F grade for packet loss - usually as much as 50% packet loss.

 

Using preemptive shows me 0% packet loss 100% of the time...

Link to comment
Share on other sites

  • Administrators

Does DumaOS have the preemptive and reactive options? Interesting to see some quality grades below A+.

 

The reason I ask is because if I use reactive or device prioritisation on the R1, DSLReports reports packet loss as shown by a quality grade other than A+. If I use both reactive and device prioritisation with share excess disabled, there's a larger amount of packet loss. If I do that and then drop my sliders right down (like

 

Using preemptive shows me 0% packet loss 100% of the time...

 

There's no algorithms to choose between. It will get you consistent grades of A at least that I have seen. We also found out that they use port 80 to do their tests which is not hardware accelerated due to Netgear's blocking feature in settings, next firmware addresses that so it can be hardware accelerated.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...