RT-AC87U, ipv6, traditional QOS

[nop0x90]

I read in the changelog that "traditional QOS is still broken" but I wasn't able to find the thread that documented why, or if the RT-AC87U is affected.

I've been using traditional QOS w/o any problems at all, until I recently enabled the IPV6 stack. In that case, IPV4 bandwidth is still what I would expect (28-29Mbit), but IPV6 seems to be capped at 5. Disabling QOS then yields 28-29Mbit on both IPV4 and IPV6.

I've switched over to Adaptive QOS, and things are working, but I've never liked this setting. It has never seemed as responsive as Traditional.

Is Traditional broken like the changelog implies, or is this IPV6+Traditional a possible bug?

 

[Undesirable]

Well, not many people were reporting about it; perhaps you're the first to confirm that traditional QoS works on IPv4 and not v6. Personally, I'm grateful for the info, because it may convince me to install the latest firmware on my RT-AC87U.

 

[Stephen Becker]

I had been using Traditional QoS for years and it randomly stopped working last week. It started limiting my download speed to the same level I had set for uploads. I literally had not even logged into my router in weeks, so didn't think it could be a settings issue. Spent hours on the phone trying convenience my useless IPS's tech support the problem was on their end, until I realized it was my own QoS (how often does that happen, lol).

Once I figured out QoS was the problem, I tried a number of things, turning it off and on, resetting settings. Tried everything expect a full factory reset of the entire router, but could not get it working. I just switched to Adaptive QoS. The techy in me doesn't like giving up control, but I have to assume Asus designed that decently well, and realistically, I have enough bandwidth that I probably don't need QoS at all.

I am on a RT-AC87R running Merlin 380.59 (updated long before problem started)

 

[Jerry Leveillee]

I don't suppose a resolution was found for this issue?? How do we get this on Merlin's radar? Seems john9527 was able to resolve this in a different fork.