[Release] FreshJR Adaptive QOS (Improvements / Custom Rules / and Inner workings)

[pattiri]

I use a lot of VPN. How is this classified? How do we make a rule for this?

I don't think this can be done. I guess router will classify your VPN traffic as "General" because when I connect to my company network VPN on my laptop, all traffic seems as "General"

 

[Wutikorn]

I don't think this can be done. I guess router will classify your VPN traffic as "General" because when I connect to my company network VPN on my laptop, all traffic seems as "General"

Most encrypted data, such as VPN traffic, are classified as "General", which means that the TM engine could not really recognize it. However, I thought the engine knows about OpenVPN traffic, not sure if you need to use default port for it to recognize or not.

 

[Vexira]

The script seems to work on the latest build. Still might need a compatibility update.

 

[FreshJR]

I've been away a little bit.

I have an older modem, AC-68U.

Looks like its EOL, with me still being on 380.68_4.

I do not plan on purchasing a new router anytime soon.

Also, I do not know if I would even consider purchasing an ASUS again. I loved this community and all of Merlin's work. Currently, from what I am reading, ASUS is slowly converting everything currently open/modifiable into closed source.

Thank's to Merlin, and the rest of the community, lot's of cool tweaks have been created, let alone all the bug fixes. Without that, the original firmware is a littered with a mess of bugs.

Being stuck with bugs we can fix ourselves was one thing, but being stuck with unfixable bugs, which ASUS has no intention of fixing, is another. I am not going to reward ASUS with a hardware sale if I am going to be stuck with a poor working product.

I may try out Dlink streamboost QOS technology next if I need a new router. Asus product support has been phenomenal and they do not owe us open source firmware, but it sure played a big part in me recommending and considering more Asus routers in the future.

EDIT: Looks like Merlin said other routers will be updated to 382 when Asus releases the code. I will look into possible fixes at that moment. I'm leaving everything on hold for now.

 

[RMerlin]

I have an older modem, AC-68U.

Looks like its EOL, with me still being on 380.68_4.

It's not. It just hasn't been migrated to the 382 code base yet - Asus only released a first GPL for it less than a week ago.

 

[JDB]

Just a small update on my VoIP classification journey! I'm now running the RT-AC5300 as my router and so am back to 380_68.4 (so the script is working nicely). I've found that one of my VoIP Clients RTP is classified correctly, but on another it goes into Other (which is better than Default at least!). Still trying to work out the differences in how the QOS is classifying the streams but some progress - at least I have a working example now!

 

[MagikMark]

Is there a way to create a “VPN” category? How? I use VPN a lot for privacy and security


Sent from my iPhone using Tapatalk

 

[FreshJR]

For VPN filtering, find something you can filter by.

-TOS/DSCP packet mark would be my first approach.
-If that doesn't work but it's always the same VPN server, use the server's IP as a filter.
-If all else fails use port filtering

You cannot create new categories, but I would just repurpose any one of the existing categories.

 

[MagikMark]

For VPN filtering, find something you can filter by.

-TOS/DSCP packet mark would be my first approach.
-If that doesn't work but it's always the same VPN server, use the server's IP as a filter.
-If all else fails use port filtering

You cannot create new categories, but I would just repurpose any one of the existing categories.

Otherwise, VPN will just go to “Default” category, right?


Sent from my iPhone using Tapatalk

 

[FreshJR]

Otherwise, VPN will just go to “Default” category, right?


Sent from my iPhone using Tapatalk

The script should have all unifentified traffic going into "others". So really nothing should be in "default".

It was last tested 380.67. If you have traffic going into "default" with the script on something is either fully or partially not working.

Try rebooting the router and observing the log for messages from adaptive QOS.

Currently all fixes are on hold. I am on an old firmware and there is no reason look into it until I am on latest.

 

[Vexira]

I've been away a little bit.

I have an older modem, AC-68U.

Looks like its EOL, with me still being on 380.68_4.

I do not plan on purchasing a new router anytime soon.

Also, I do not know if I would even consider purchasing an ASUS again. I loved this community and all of Merlin's work. Currently, from what I am reading, ASUS is slowly converting everything currently open/modifiable into closed source.

Thank's to Merlin, and the rest of the community, lot's of cool tweaks have been created, let alone all the bug fixes. Without that, the original firmware is a littered with a mess of bugs.

Being stuck with bugs we can fix ourselves was one thing, but being stuck with unfixable bugs, which ASUS has no intention of fixing, is another. I am not going to reward ASUS with a hardware sale if I am going to be stuck with a poor working product.

I may try out Dlink streamboost QOS technology next if I need a new router. Asus product support has been phenomenal and they do not owe us open source firmware, but it sure played a big part in me recommending and considering more Asus routers in the future.

EDIT: Looks like Merlin said other routers will be updated to 382 when Asus releases the code. I will look into possible fixes at that moment. I'm leaving everything on hold for now.

Thank you for coming back.

 

[Xruptor]

Also, I do not know if I would even consider purchasing an ASUS again. I loved this community and all of Merlin's work. Currently, from what I am reading, ASUS is slowly converting everything currently open/modifiable into closed source.

Thought it was only a few features and or third-party stuff that was closed source. Other then that I was under the assumption everything else is still pretty much open-source. I highly doubt Asus will convert everything to closed-source. They will lose a lot of fan followers for their routers if they did this. Many of the reasons people flock to ASUS routers is their flexbility and availability for tinkering.

Is the other companies as closed source as Asus in that aspect then? I'm not entirely sure about Netgear, D-Link, Linksys, etc... I'm sure they all have their fair share of closed source features.

 

[Vexira]

Thought it was only a few features and or third-party stuff that was closed source. Other then that I was under the assumption everything else is still pretty much open-source. I highly doubt Asus will convert everything to closed-source. They will lose a lot of fan followers for their routers if they did this. Many of the reasons people flock to ASUS routers is their flexbility and availability for tinkering.

Is the other companies as closed source as Asus in that aspect then? I'm not entirely sure about Netgear, D-Link, Linksys, etc... I'm sure they all have their fair share of closed source features.

They already started close sourceing components, because of people like xwrx vortex, who have been porting the firmware to devices that are of asus competitors.

 

[RMerlin]

Things are worse with most of Asus's competitors. Netgear even closes the source code to their iptables configuration...

 

[Vexira]

Any updates considering a 68u build exists?

 

[brummygit]

Looks like Merlin said other routers will be updated to 382 when Asus releases the code. I will look into possible fixes at that moment. I'm leaving everything on hold for now.

Hi @FreshJR, RMerlin released his 382.2 Beta firmware today for the RT-AC68U and it seems to be pretty good. Your script made a huge difference to my network on previous releases so thanks for your past hard work, I would love to continue to run it on my router if you have the time and enthusiasm for further development.

 

[deddc23efb]

Thought it was only a few features and or third-party stuff that was closed source. Other then that I was under the assumption everything else is still pretty much open-source. I highly doubt Asus will convert everything to closed-source. They will lose a lot of fan followers for their routers if they did this. Many of the reasons people flock to ASUS routers is their flexbility and availability for tinkering.

Is the other companies as closed source as Asus in that aspect then? I'm not entirely sure about Netgear, D-Link, Linksys, etc... I'm sure they all have their fair share of closed source features.

I'm thinking along the lines of freshJR too. I had an aging RT-N66 running Shibby that worked great. I thought I'd be safe buying an updated system, get a speed boost and try out the TCP Codel updates ... but the other alternatiove firmwares have fallen by the wayside, ASUS ships firmware with gaping bugs and closed source major features (the Adaptive QOS is a disaster ... ).

I had considered building a pfSense box and thought a new RT-AC3100 would be simpler. It wasn't. The fact that AB-Solution/Skynet and other CLI only features exist and completely broken features like ASUS's terrible Network Summary or busted traditional QOS is shocking. I won't recommend ASUS products.

I'm trying to get my system limping along with some semblance of manageable QOS and leave it alone as good enough. It looks like the consumer router world is in a race to the bottom.

 

[Gitsum]

I'm thinking along the lines of freshJR too. I had an aging RT-N66 running Shibby that worked great. I thought I'd be safe buying an updated system, get a speed boost and try out the TCP Codel updates ... but the other alternatiove firmwares have fallen by the wayside, ASUS ships firmware with gaping bugs and closed source major features (the Adaptive QOS is a disaster ... ).

I had considered building a pfSense box and thought a new RT-AC3100 would be simpler. It wasn't. The fact that AB-Solution/Skynet and other CLI only features exist and completely broken features like ASUS's terrible Network Summary or busted traditional QOS is shocking. I won't recommend ASUS products.

I'm trying to get my system limping along with some semblance of manageable QOS and leave it alone as good enough. It looks like the consumer router world is in a race to the bottom.

Have you looked at Untangle?

 

[deddc23efb]

Have you looked at Untangle?

Thanks I didn't know about Untangle. I'll check it out.

 

[deddc23efb]

The script should have all unifentified traffic going into "others". So really nothing should be in "default".

It was last tested 380.67. If you have traffic going into "default" with the script on something is either fully or partially not working.

Try rebooting the router and observing the log for messages from adaptive QOS.

Currently all fixes are on hold. I am on an old firmware and there is no reason look into it until I am on latest.

Hi FreshJR,

I looked through your script to see what might be broken. It looks like the match strings are slightly different in 382.. The second field has changed from 0x8000FFFF to 0xC000FFFF, as an example. In general all the of the second field values have a 4 OR'd in (0x8 OR 0x4 == 0xC) There are also now 6 or 7 different entries in the TC filters that detect the FFFF fields. On my system only the first one is used - not sure what triggers the Trend Micro code to use the other fields ...perhaps it's prio.

It also looks at first glance that ASUS has cleaned up their rate/ceil allocations a little bit. A quick glance at the current values in the tc class dump for 382 shows values that look pretty much like what your default math in the script does. I haven't stared at that so much so perhaps I've missed a detail, but the ASUS defaults might be ok.

I made some minor mods to your script to look for the first instance of the "Unidentified" Filter and ignore the rest as well as search for 0xC.... fields. It appears to do the right thing on my system. Generalizing these changes to support both 380 and 382 plus truly understanding what the other entries do would be some more work.

All that to say it's probably not too much work to get your script going for the 382 people again.

/dedd