[alpha] Early preview builds for Asuswrt-Merlin 380.66

[Naftali Oziel]

Wifi is a complex technology, where timings can be critical. Doesn't take much to create compatibility issues with other devices. MU-MIMO, Airtime Fairness, Beamforming - those are all extensions that haven't been fully tested by everyone involved (i.e. both the router AND the client manufacturers).

If am in N only mode, is there a purpose for airtime fairness to be enabled?

 

[Skywalker]

Running for a couple of days on my RT-AC66U now, Wifi 5G seems to be more stable then on 380.65.

 

[DaveMishSr]

If am in N only mode, is there a purpose for airtime fairness to be enabled?

Airtime fairness is mostly used when you have slower legacy devices paired with newer, faster devices. These devices need more resources to accomplish similar workloads so they take away resources from faster network clients. So enabling airtime fairness decreases the performance of faster devices when you have them connected with legacy devices on the same network.

Edit: I originally misspoke here and said increases instead of decreases. Sorry for the confusion.

You should be able to see where this is going. By selecting N only you have narrowed the network protocols to one so you have eliminated the slower network protocols. Therefore, airtime fairness becomes irrelevant. While it shouldn't affect your network performance if it were to be enabled, I am always of the mind that less is more. In this specific instance I would not enable it.

I should have also included:
1) long physical distance
2) weak signal strength,

as being contributing factors to slow the speed of wifi devices.

Here is a link which better explains airtime fairness. I strongly encourage you to read it in its entirety.

https://routerguide.net/airtime-fairness-on-or-off/

 

[stambeccuccio]

..By selecting N only you have narrowed the network protocols to one so you have eliminated the slower network protocols. Therefore, airtime fairness becomes irrelevant...

...not exactly:
The relatively “slow” wifi speed devices can be slow from:
1) long physical distance
2) weak signal strength,
3) or simply being a legacy device with older technology.

 

[Reinvented]

Here to report, I've put a load on my router and it's still doing just fine. Connections are holding up, and the QoS is classifying traffic correctly.

However, I do have one problem that I've had since 380.65 and that's the Traffic Analyzer isn't showing any incoming traffic on the 5 GHz or 2.4 GHz network tabs (the blue colored), which is pretty strange still. Is there something to remedy this?

 

[WET]

I would like to report the following:
When I flashed my RT-AC68U to 380.66 my wired HP Printer became unavailable for some clients. When I disabled airtime fairness the printer became available for the clients that had lost the printer but my two thermostats which are G clients would no longer hook up to router when airtime fairness was disabled.

I have flashed back to 380.65

I have now determined that one client of mine had a print spooler problem which I have now corrected. I have had no issue with my wired HP Printer on 380.65, 380.65-2 or 380.66 alpha2 which I am now running on.

 

[quant88]

Flashed 380.66 to my RT-AC88U. Basic wifi functions are working ok. Not using QOS and VPN features.
However, I am unable to print on the Samsung wireless printer with this alpha version. Revert back to 380.65_2 and it is wifi printing is fine.

 

[RMerlin]

Flashed 380.66 to my RT-AC88U. Basic wifi functions are working ok. Not using QOS and VPN features.
However, I am unable to print on the Samsung wireless printer with this alpha version. Revert back to 380.65_2 and it is wifi printing is fine.

Have you read the included Changelog?

 

[Blinkyz]

Changelog
Known issues:
* Some wifi clients such as wireless printers
have connectivity issues. Try disabling
Airtime Fairness as a potential workaround.

 

[DaveMishSr]

...not exactly:
The relatively “slow” wifi speed devices can be slow from:
1) long physical distance
2) weak signal strength,
3) or simply being a legacy device with older technology.

You are correct. However, I posted "mostly used", specifically spoke to item #3 and included a link that went over airtime fairness in much greater detail. Thanks for your clarification though.

 

[Radiohead]

Stable after 7 days running. http://prntscr.com/elwf0v

The only issue is after a reboot I need to stop and restart the secondary WAN connection before it goes to standby. But then this has been the case with previous versions as well. Is this default behaviour? Fail over and fail back work well with a FTTx and a DSL connection.

 

[wsarahan]

So far this build have way more stable Wifi over the latest Asus FW 7266

Just a question, by default the Airtime Fairness is enabled, i have no idea what this thing does, just read here inthe topic that some people are disabling it because of printes....

I don`t have a printer in my enviroment, should i let this on or off? Mimo is on as well in both 3 bands

Thanks for the nice work again @RMerlin

 

[kibosh]

You are using 380.65 which is based on an older GPL. Asus broke airtime fairness sometime after, that is why you don't have an issue.

 

[wsarahan]

You are using 380.65 which is based on an older GPL. Asus broke airtime fairness sometime after, that is why you don't have an issue.

So should i let this one or or off in this alpha2 build?

 

[bluepoint]

So should i let this one or or off in this alpha2 build?

Test it yourself. Airtime fairness seems to have an effect with my two sons iPhone 7/7+ that are both at the edge of the wireless range(-70dBm) with intel modems(AT&T). Both are complaining lags when streaming video when "airtime fairness" is enabled. Disabling AF solves that problem. Ill see hopefully, when iOS 10.3 is released next week possibly new intel driver will be introduced and solves the lagging.

 

[Reinvented]

Test it yourself. Airtime fairness seems to have an effect with my two sons iPhone 7/7+ that are both at the edge of the wireless range(-70dBm) with intel modems(AT&T). Both are complaining lags when streaming video when "airtime fairness" is enabled. Disabling AF solves that problem. Ill see hopefully, when iOS 10.3 is released next week possibly new intel driver will be introduced and solves the lagging.

This lag your sons experiences is one of the effects of airtime fairness. Probably because they are so far away, and therefore it doesn't prioritize those devices at all. In fact, it will reduce the amount of traffic that they receive like you noticed, and immediately were able to stream again when it was turned off.

In my opinion, it's a useless feature and causes more problems than not. Proper QoS setup and proper wireless settings will trump airtime fairness any time.

 

[kibosh]

So should i let this one or or off in this alpha2 build?

I have not noticed any adverse effects in disabling airtime fairness. As @RMerlin said earlier, "MU-MIMO, Airtime Fairness, Beamforming - those are all extensions that haven't been fully tested by everyone involved." I can tell you it definitely fixes airprint issues and also with my wireless clients disappearing and reappearing at random in the client list.

 

[bluepoint]

This lag your sons experiences is one of the effects of airtime fairness. Probably because they are so far away, and therefore it doesn't prioritize those devices at all.

I was thinking about that but when I tried my wife's iPhone 6s that has QUALCOMM modem, the lagging is not there in the same areas. Most probably intel needs adjustments since these modems were the first ones with these models. I've read also they are slower than the QUALCOMM's but I didn't test it to bother with my sons.

 

[Reinvented]

I was thinking about that but when I tried my wife's iPhone 6s that has QUALCOMM modem, the lagging is not there in the same areas. Most probably intel needs adjustments since these modems were the first ones with these models. I've read also they are slower than the QUALCOMM's but I didn't test it to bother with my sons.

Not sure why you keep thinking it's a driver issue. It's clearly not...

 

[bluepoint]

Not sure why you keep thinking it's a driver issue. It's clearly not...

No argument needed, I stay with what I think it is, two different hardwares different results.