FlexQoS FlexQoS 1.0 - Flexible QoS Enhancement Script for Adaptive QoS

[lilstone87]

I have


I have an AX88U and Wi-Fi traffic is showing fine on the upstream and downstream in qos with the new script. Perhaps there is something funky going on with your router and it requires a factory reset?

I wish that was the case, however I wasn't using my AX88U for a bit till yesterday. Which yesterday, I flashed RMerlin's latest beta on it, which is 384.18_beta1. After I factory reset the router, before manually setting it back up.

As I said in my other posts it's only upstream data over wifi I noticed this happening on. For me, it was while running the speedtest app on my phone.

To be honest.. I happen to browse the asus official sub forum on here, before I fell asleep this morning. Which is where everyone post Asus latest firmwares for their routers. Low, and behold, a recent firmware posted for one of their AX router. There was a comment, or two, of people complaining their upload traffic isn't being passed thru Adaptive QoS still. As I mentioned in my other post, I feel this has been an ongoing issue for a few months now, and Asus still hasn't fixed it.

Edit: Here's the post of the other person having the same issue on their AX router. I have a feeling this might be a wifi driver related issue with their AX routers.
RT-AX58U Firmware 3.0.0.4.384_9354 -- 6/11/2020

 

[danielk]

Simplest is to make sure jffs isn’t full. If not, can you post /jffs/addons/flexqos/table/flexqos_arrays.js to pastebin.com? And make sure you have .db files in /tmp/bwdpi/

Thanks, but i dont know how to do any of that If you feel like getting specific on how to do all that i'd appreciate it! However i also completely understand if guiding simpletons like myself isnt what you signed up for

 

[dave14305]

Thanks, but i dont know how to do any of that If you feel like getting specific on how to do all that i'd appreciate it! However i also completely understand if guiding simpletons like myself isnt what you signed up for

In SSH:

df /jffs

Or look on the Tools page in the GUI.

 

[danielk]

In SSH:

df /jffs

Or look on the Tools page in the GUI.

df /jffs
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/mtdblock4 64256 3180 61076 5% /jffs
danielk@RT-AC68U-54F8:/tmp/home/root#

 

[dave14305]

df /jffs
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/mtdblock4 64256 3180 61076 5% /jffs
danielk@RT-AC68U-54F8:/tmp/home/root#

How about:

ls -l /tmp/bwdpi/bw*.db

 

[lilstone87]

@dave14305

I know my current issue with my wifi upstream data not being controlled within Adaptive QoS, doesn't have anything to do with your updated script. However I have been thinking a bit again, early this morning. I am pretty sure a few months back, I stopped using my AX88U due to me, and others, noticing upload data on wifi devices not being run thru Adaptive QoS.

With that, as I said in my other post. I put my AX88U back online yesterday, flashed RMerlin's latest beta build on it, factory reset, and manually set it up. I am one of the people who believe in factory resetting often, to minimize any slight chance of a weird bug slipping threw causing an issue. So today, I am gonna try to factory reset my AX88U again, and try running Adaptive QoS first without the script, and test my phone running the speedtest app. I have a good feeling the issue will appear once again, and no further testing will be needed.

I know the user a couple post above said he isn't having this issue. Which imho I will just call him lucky. As I know this was an issue for me a few months ago, and the reason I stopped using my AX88U, and went back to using my older RT-AC3100. Also I now remember when I noticed this before, others reported the same issue as well.

With all of this said, If @RMerlin has the time to reply to this post of mine, in the coming days. I would greatly appreciate him commenting on this. As I have a feeling he has seen this talked about in the recent past as well. Currently I am gonna stick with my thought of something within the wifi driver of current AX routers, is broken, causing upstream data from wifi devices to not pass thru Adaptive QoS.

 

[danielk]

How about:

ls -l /tmp/bwdpi/bw*.db

Spoiler: result

danielk@RT-AC68U-54F8:/tmp/home/root# ls -l /tmp/bwdpi/bw*.db
-rw-rw-rw- 1 danielk root 33211 Jun 27 01:20 /tmp/bwdpi/bwdpi.app.db
-rw-rw-rw- 1 danielk root 17700 Jun 27 01:20 /tmp/bwdpi/bwdpi.appdb.db
-rw-rw-rw- 1 danielk root 73 Jun 27 01:20 /tmp/bwdpi/bwdpi.beh.db
-rw-rw-rw- 1 danielk root 570 Jun 27 01:20 /tmp/bwdpi/bwdpi.bndwth.db
-rw-rw-rw- 1 danielk root 487 Jun 27 01:20 /tmp/bwdpi/bwdpi.cat.db
-rw-rw-rw- 1 danielk root 285351 Jun 27 01:20 /tmp/bwdpi/bwdpi.devdb.db
-rw-rw-rw- 1 danielk root 398575 Jun 27 01:20 /tmp/bwdpi/bwdpi.rule.db
danielk@RT-AC68U-54F8:/tmp/home/root#

 

[dave14305]

Spoiler: result

danielk@RT-AC68U-54F8:/tmp/home/root# ls -l /tmp/bwdpi/bw*.db
-rw-rw-rw- 1 danielk root 33211 Jun 27 01:20 /tmp/bwdpi/bwdpi.app.db
-rw-rw-rw- 1 danielk root 17700 Jun 27 01:20 /tmp/bwdpi/bwdpi.appdb.db
-rw-rw-rw- 1 danielk root 73 Jun 27 01:20 /tmp/bwdpi/bwdpi.beh.db
-rw-rw-rw- 1 danielk root 570 Jun 27 01:20 /tmp/bwdpi/bwdpi.bndwth.db
-rw-rw-rw- 1 danielk root 487 Jun 27 01:20 /tmp/bwdpi/bwdpi.cat.db
-rw-rw-rw- 1 danielk root 285351 Jun 27 01:20 /tmp/bwdpi/bwdpi.devdb.db
-rw-rw-rw- 1 danielk root 398575 Jun 27 01:20 /tmp/bwdpi/bwdpi.rule.db
danielk@RT-AC68U-54F8:/tmp/home/root#

cut -b -80 /jffs/addons/flexqos/table/flexqos_arrays.js

Also, just make sure any Adblockers are disabled for the router URL.

 

[danielk]

cut -b -80 /jffs/addons/flexqos/table/flexqos_arrays.js

Also, just make sure any Adblockers are disabled for the router URL.

danielk@RT-AC68U-54F8:/tmp/home/root# cut -b -80 /jffs/addons/flexqos/table/flexqos_arrays.js
var catdb_mark_array = [ "000000", var catdb_label_array = [ "Untracked",
danielk@RT-AC68U-54F8:/tmp/home/root#

And my microsoft Edge is completely unmodded (unused..) so should be no interference from adblockers or anything like that

Edit: same results go for "clean" firefox, even in incognito mode.

 

[dave14305]

danielk@RT-AC68U-54F8:/tmp/home/root# cut -b -80 /jffs/addons/flexqos/table/flexqos_arrays.js
var catdb_mark_array = [ "000000", var catdb_label_array = [ "Untracked",
danielk@RT-AC68U-54F8:/tmp/home/root#

And my microsoft Edge is completely unmodded (unused..) so should be no interference from adblockers or anything like that

Edit: same results go for "clean" firefox, even in incognito mode.

Ok, probably a timing issue if you rebooted your router recently. Run:

rm /jffs/addons/flexqos/table/flexqos_arrays.js
service restart_firewall

and refresh the page after a short wait.

 

[dave14305]

Currently I am gonna stick with my thought of something within the wifi driver of current AX routers, is broken, causing upstream data from wifi devices to not pass thru Adaptive QoS

I noticed a recent commit in the ax branch where the AX58U interface for tc WAN traffic was eth4, where we have always assumed eth0. So just my way of saying the sands are shifting and old assumptions may no longer be valid. We will overcome!

 

[lilstone87]

@Phil Outram

I first just want to thank you for trying to help. However I had already did what you suggested yesterday, before I posted about this problem. Overall I am one of the ones who often factory reset their routers after most firmware updates, and such.

I want to mention again, I am running RMerlin's latest beta build on my AX88U. Which is 384.18_beta1. I hope you are running this build as well. Being that I woke up early this morning, and no one in my house is really awake yet. I decided to once again do another factory reset. I did the initial router setup, which includes setting up the wifi. After that was done. I simply went, and enabled Adaptive QoS, used FQ_Codel, I did not set any overhead value. I set my download at 500mbps, and my upload at 24mbps. After that I hit apply to start QoS. I gave that a few moments, and pulled up my web browser. Loaded up fast.com, and it gave me near 500mbps down, and a little under 24mbps up. I also loaded up speedtest.net, pick a local server, ran another test, I got 480+ down, and 23 up.

Overall both tests on my wired pc, gave me results right under the limits I set. So after that, I pulled my phone up, reconnected to the wifi. I made sure to forget the network during the factory reset as well. I opened up the speedtest app on my phone. Ran a test to the same server, that I ran on my wired pc. I ended up running two test on my phone. Both in the 34-35mbps range. Which is my max upload from my provider. I also made sure to have the pie chart tab up on the router, while I tested. During the download, data showed on the download pie chart, with active data being passed. When the upload started.. nothing showed up once again.

Finally.. as I write this, I thought about my wired pc, which I do have wifi on. So I disconnected ethernet, and connected via wifi to the AX88U. Ran another test on speedtest.net, to the same server. On the upload, once again upload was maxed out at around 35mbps. Also the pie chart page once again showed no active data while the upload part of the test ran. Overall all this testing on my end, proves something isn't right when it comes to the wifi/wifi driver on this AX88U. As wifi upstream data bypasses Adaptive QoS.

 

[danielk]

Ok, probably a timing issue if you rebooted your router recently. Run:

rm /jffs/addons/flexqos/table/flexqos_arrays.js
service restart_firewall

and refresh the page after a short wait.

I rebooted the entire router first, and still no content. Then i tried the rm command and restarted, refreshed and that seems to have fixed it!

Thank you for the hand-holding!

Now i just need dslreports.com to come back online so i can do some bloatbuffer testing..

 

[lilstone87]

I noticed a recent commit in the ax branch where the AX58U interface for tc WAN traffic was eth4, where we have always assumed eth0. So just my way of saying the sands are shifting and old assumptions may no longer be valid. We will overcome!

So is Asus the one to blame for this still? I am a bit confused on this post by you. I just posted more on me testing this, this morning. I didn't see this post of yours before I posted it. Overall I hope this can be corrected in the near future. As this to me.. is a deal breaker, as Adaptive QoS is broken in this state imho. Especially when someone like me who has a lower upstream limit, that can easily be hit, via a wifi device. Causing some higher than expected latency spikes.

Anyways no more of this talk from me, for a bit now. I think I posted enough info about it, in your thread. Which overall isn't a problem with your script at all. So hopefully you don't mind me discussing what I did here. I just want this to be fixed, with Adaptive QoS able to run at the best of its ability.

EDIT: Let me be clear.. when I say Adaptive QoS is broken in its current state. Overall I don't blame the feature itself. It's just current upstream wifi data is bypassing it, and so overall, it isn't able to do its job correctly.

 

[dave14305]

I rebooted the entire router first, and still no content. Then i tried the rm command and restarted, refreshed and that seems to have fixed it!

Thank you for the hand-holding!

Now i just need dslreports.com to come back online so i can do some bloatbuffer testing..

Thanks for finding my bug.

 

[Vexira]

So is Asus the one to blame for this still? I am a bit confused on this post by you. I just posted more on me testing this, this morning. I didn't see this post of yours before I posted it. Overall I hope this can be corrected in the near future. As this to me.. is a deal breaker, as Adaptive QoS is broken in this state imho. Especially when someone like me who has a lower upstream limit, that can easily be hit, via a wifi device. Causing some higher than expected latency spikes.

Anyways no more of this talk from me, for a bit now. I think I posted enough info about it, in your thread. Which overall isn't a problem with your script at all. So hopefully you don't mind me discussing what I did here. I just want this to be fixed, with Adaptive QoS able to run at the best of its ability.

EDIT: Let me be clear.. when I say Adaptive QoS is broken in its current state. Overall I don't blame the feature itself. It's just current upstream wifi data is bypassing it, and so overall, it isn't able to do its job correctly.

Trend micro would be to balme more or less since they make the module

 

[lilstone87]

Trend micro would be to balme more or less since they make the module

Thanks for the info. I just hope someone can get this info to Asus, and whoever can fix it, fixes it. As this has been an ongoing issue for a few months now. I stopped using the AX88U before, because of it. Which sadly.. I forgot about it, when I seen the recent Cake/FlexQoS work being done. It got me excited, but once again, I quickly noticed this issue. As I am one who does a bit of testing, and this issue easily sticks out, when testing wifi upstream data.

 

[ugandy]

EDIT: Let me be clear.. when I say Adaptive QoS is broken in its current state. Overall I don't blame the feature itself. It's just current upstream wifi data is bypassing it, and so overall, it isn't able to do its job correctly.

do i read you correctly? adaptiveQoS works only on the downstream due to a bug?

 

[lilstone87]

do i read you correctly? adaptiveQoS works only on the downstream due to a bug?

No you're not reading that correctly. On my AX88U, and likely other Asus AX model routers. Upstream data OVER wifi, isn't being passed thru AdaptiveQoS correctly right now. Wired to the router, is working just fine.

 

[ugandy]

No you're not reading that correctly. On my AX88U, and likely other Asus AX model routers. Upstream data OVER wifi, isn't being passed thru AdaptiveQoS correctly right now. Wired to the router, is working just fine.

ok, so only wifi data. still bad. it would explain why i got no latency control on the UL, when i tested flexqos, while saturating the UL via wifi. that's no go for me on flexqos then :-(