FlexQoS FlexQoS 1.0 - Flexible QoS Enhancement Script for Adaptive QoS

[Phil Outram]

Version 0.8.0

• NEW: Backported Merlin connection table filters (kept FreshJR local device filter dropdown)
• NEW: Added connection table counts (total and shown/unfiltered)
• NEW: Backported Merlin QoS class tooltip on Application name
• FIXED: personalized URL after install/update/help (@maghuro)
• FIXED: cleanup extra html tags
• FIXED: Device Filter menu spacing

sh /jffs/addons/flexqos/flexqos.sh update

Unfortunately the changes to the chart page seems to have broken the auto refresh, it is now disabled due to too many tracked connections.

 

[QuikSilver]

Version 0.8.0

• NEW: Backported Merlin connection table filters (kept FreshJR local device filter dropdown)
• NEW: Added connection table counts (total and shown/unfiltered)
• NEW: Backported Merlin QoS class tooltip on Application name
• FIXED: personalized URL after install/update/help (@maghuro)
• FIXED: cleanup extra html tags
• FIXED: Device Filter menu spacing

sh /jffs/addons/flexqos/flexqos.sh update

More cosmetic than anything...Would renaming the Local IP/Remote IP section to something like "Outgoing/Incoming" be better suited for the table filter? Filtering by "Local IP" filters by "Remote IP" as well. Not sure if intended or not. Name resolution doesn't seem to be working on the Remote IP section either.

 

[QuikSilver]

More cosmetic than anything...Would renaming the Local IP/Remote IP section to something like "Outgoing/Incoming" be better suited for the table filter? Filtering by "Local IP" filters by "Remote IP" as well. Not sure if intended or not. Name resolution doesn't seem to be working on the Remote IP section either.

Forgot to add picture for example of filter by Local IP

 

[Jack Yaz]

Version 0.8.0

• NEW: Backported Merlin connection table filters (kept FreshJR local device filter dropdown)
• NEW: Added connection table counts (total and shown/unfiltered)
• NEW: Backported Merlin QoS class tooltip on Application name
• FIXED: personalized URL after install/update/help (@maghuro)
• FIXED: cleanup extra html tags
• FIXED: Device Filter menu spacing

sh /jffs/addons/flexqos/flexqos.sh update

in hindsight i should have used tableapi for uiDivStats' query log - but it was fun writing my own implementation

 

[dave14305]

Unfortunately the changes to the chart page seems to have broken the auto refresh, it is now disabled due to too many tracked connections.

Fair enough. I never reached more than 750 connections on my own network to see how it behaved in real life. I just lifted it straight from Merlin's classification page. I've reverted that particular change for now with v0.8.1. Check for an update.

I'd like to think I can add this back using a session cookie to disable it on initial load, but let the user override it for the rest of the session. I'll need to plagiarize some of @Jack Yaz cookie code.

 

[dave14305]

View attachment 24377

Forgot to add picture for example of filter by Local IP

Is this LAN to WAN connection or WAN to LAN? The old FreshJR code is supposed to switch these, but maybe I broke something when adding the filtering. Does your DHCP range start with 192.168.1.? Just trying to imagine what's gone wrong without really looking yet.

 

[dave14305]

in hindsight i should have used tableapi for uiDivStats' query log - but it was fun writing my own implementation

Everything I read says it's slow for large tables. Neither FreshJR nor Merlin ended up using it for their connection lists.

 

[dave14305]

View attachment 24377

Forgot to add picture for example of filter by Local IP

I think I broke it when I tried to also detect WAN to LAN IPv6 connections and swap them, but since I currently have IPv6 enabled, I didn't test what would happen when the IPv6 prefix was blank (i.e. IPv6 disabled). So I'm guessing that it's swapping everything. I reverted that change for now. Pushed v0.8.2.

Somebody here lacks coding discipline...name rhymes with save14305...

 

[dave14305]

Version 0.8.3 pushed to fix the swapping of local/remote IPs with the new filtering logic.

The real reason for these bugs is that I really want to learn how to use the git command line. This gives me lots of practice.

If anyone has real WAN to LAN connections that do not show-up correctly in the list, please let me know and send a screenshot. I don't know how likely this scenario is, but since it was part of FreshJR_QOS, I've tried to leave it as-is even though I have no use for it personally in my own network.

 

[QuikSilver]

I think I broke it when I tried to also detect WAN to LAN IPv6 connections and swap them, but since I currently have IPv6 enabled, I didn't test what would happen when the IPv6 prefix was blank (i.e. IPv6 disabled). So I'm guessing that it's swapping everything. I reverted that change for now. Pushed v0.8.2.

Somebody here lacks coding discipline...name rhymes with save14305...

Looks correct now. Thanks! Don't be so hard on yourself.

“Congratulate yourself when you fail - you're learning.”
― Stewart Stafford

 

[Sinner]

Thanks for taking a look! The AC68U isnt really running anything, only thing ive manually installed is FlexQOS, apart from that I've disabled most functions, even the 5ghz wireless band. And for context: i wasnt even using QoS until a few days ago when i got a new gateway from my ISP.

Games are recognized and categorized correctly in Flex (at max prio after net control, followd by voip) but I havent had time to test much. But previously i used to have very random spurs of issues where some days were perfectly fine (like most of my test in previous post), then others where id suddenly be getting weird latency issues for anywhere between 5-60 seconds. Symptoms would be laggy voip (discord) and banding or teleporting enemies in games - or in games where available ping statistics would suddenly show red lines indicating packet loss and +100 increase in ms before setling back to 40-50 ish standards. This would occur even when i was the only connected device online. Also during the dslreports.com tests i had quite a few of them spike the bufferbloat gauge to a red 500+ ms just for a split second then return to a green value, but my final grades turned out decent non the less..

Though, i do often have a lot of traffic on the wireless, both cellphones and a laptop that the missus uses for streaming. Am i correct in understanding previous comments here that these devices wont be throttled correctly by the QoS? If so are there any other suggestions on how to prioritize traffic from a specific device? or de-prioiritize ALL wireless traffic vs. ethernet connected traffic? (afaik the device priority stuff under "Bandwidth Monitor" doesnt work)

qos does throttle wifi but maybe not perfectly due to wifi's own buffering system. As for your random spikes its prolly important to determine if its due to local trafic on your network or possibly general internet lags during covid19 and everyone online all the time. It also could be interference and some of yuor wifi devices reconnecting even. I have a tv in the house here that nearly hangs the router for a second everytime its turned on and connects. It can be many thing causing it but im gonna guess if its not traffic in your home causing it when its happening then its prolly NOT qos or flexqos causing it.

 

[chris.at]

One minor thing Dave:
When you filter the list by device and for example have one device with ip address .2, one with .21 and one with .215 then you see alle three devices in the list if you select .2 and with .21 selected you see this one plus .215 - looks like the filter doesn't use "equals" but "contains" for filtering.

 

[dave14305]

One minor thing Dave:
When you filter the list by device and for example have one device with ip address .2, one with .21 and one with .215 then you see alle three devices in the list if you select .2 and with .21 selected you see this one plus .215 - looks like the filter doesn't use "equals" but "contains" for filtering.

That is a good point. Most of the filters should be "contains" but the local IP should be a match. I will fix it.

 

[dave14305]

One minor thing Dave:
When you filter the list by device and for example have one device with ip address .2, one with .21 and one with .215 then you see alle three devices in the list if you select .2 and with .21 selected you see this one plus .215 - looks like the filter doesn't use "equals" but "contains" for filtering.

I pushed a fix but you'll need to force an update since I didn't change the main script version.

On the plus side, I learned to cherry-pick a commit from one branch to another.

 

[chris.at]

Wow, that was fast!
I can confirm that the filter is working now, thank you Dave!

 

[dave14305]

Wow, that was fast!
I can confirm that the filter is working now, thank you Dave!

Don't be so happy yet, I just found a new bug from the last hotfix. Force update again. Wasn't a deal-breaker but was sloppy. Then you can be happy.

 

[Sinner]

More cosmetic than anything...Would renaming the Local IP/Remote IP section to something like "Outgoing/Incoming" be better suited for the table filter? Filtering by "Local IP" filters by "Remote IP" as well. Not sure if intended or not. Name resolution doesn't seem to be working on the Remote IP section either.

That wuld be misleading tho as not all of those connections are two way. local/remote is more accurate and doesnt make any assumptions imo

 

[QuikSilver]

That wuld be misleading tho as not all of those connections are two way. local/remote is more accurate and doesnt make any assumptions imo

Based on what I was seeing it looked to be two way but it has now been fixed based on Dave's change in

Version 0.8.3 pushed to fix the swapping of local/remote IPs with the new filtering logic.

 

[JohnD5000]

I would use your lowest speedtest and apply the 95% to that value. If you still see bad bloat reduce it to 90%. Keep reducing the bandwidth until you get good results (however you like to measure it).

How do you measure bad bloat? Is that just speedtest?

 

[dave14305]

How do you measure bad bloat? Is that just speedtest?

I measure it by how hard it is to button my pants...oh, you mean bufferbloat...

I'm not too fanatical myself about bufferbloat. I run the occasional dslreports speedtest, but mostly if I notice a MS Teams or Skype call breaking up a lot during the workday, I revisit my bandwidth settings. I'm not a gamer, so I don't have any games or apps telling me I have bad ping times, etc.

As everyone will soon discover (if they haven't already)...I'm not really a QoS expert at all. I've just studied the living daylights out of the FreshJR_QOS script.