I've been following along with testing and tuning in this thread--specifically beginning with post:
See also:
1 Like
All stable here with a snapshot from a few days ago.
When you say "new firmware", are you referring to stock gl.inet firmware?
The talk is about new firmware for the Wi-Fi chipset released by Mediatek.
It's not included in the snapshots yet, but people are testing it with custom builds
or manually replacing the firmware files.
Ah - yes - thank you for the clarity. I too have manually replaced the .bin files in my build with the updates... so the temps I showed are off the new firmware. No issues here.
EDIT - to add clarity, I'm a fairly vanilla build. Kernel 6.1. 1gig down / 35up cable connection. irqbalance is enabled. packet steering is disabled. SFO & HFO enabled. WED enabled. SQM disabled.
3 Likes
So I used the latest snapshot. I am still having the same issues with the internet not working when I’m bouncing around multiple APs around the house, even though the AP shows I’m connected with a great signal. I have a MT6000 as a main router and 3 MT6000 as dumb APs. It seems something broke with either Bridger or WED. I remove bridger from the dumb APs and turned WED off on all units, and everything works perfectly. Fast roaming works great, and I don’t loose internet connectivity when I’m bouncing around the house. Been up 24 hours and no issues, but with Bridger installed on the dumb APs and WED enabled on all the units… it lasts 5-10 mins before I lose internet walking around the house.
2 Likes
You are probably right. Something is probably off in bridger. It gave me trouble with either WED on or off.
I gave up on it. Don't really need either. Using the MT6000 as (expensive) access points. CPU utilization is so low even while we clobber NAS and internet that it makes no sense to risk instability for it. I would like to use it just because it's there but I'll wait until the dust settles.
And that would really be my humble suggestion... unless you really need it, just turn it all off.
2 Likes
Re: bridger, you will probably have to open an issue, and ping the author, to get some traction on a resolution. Might need some help to firefight the issue.
2 Likes
How would I open a case? Sorry, I’m relatively new to OpenWRT. Thanks!
For bridger:
@gssjshark and @apccv Which snapshot build are you running at this point?
Using snapshots built by OpenWrt.
r26060 with bridger - 2024.04.22~40b1c5b6 and
r25938 with bridger - 2024.04.15~a08e51e6
(I'll update the old one maybe next week)
Kernel 6.1.86 on all. Default radio firmware (not replaced with latest bin blobs) on all.
I am using r26047on main router, and r26060 on my dumb APs and was using Bridger 2024.04.22~40b1c5b6. Kernel 6.1.86 with the default firmware on all MT600.
Thank you! I created the new issue here: https://github.com/nbd168/bridger/issues/6
@apccv please feel free to add anything
1 Like
After two weeks of trial and error, I decided to disable WED (and of course disable bridger). It is now operating wonderfully and stably.
2 Likes
Would you mind explaining your setup? The reason I ask, is because Bridger seems to be the culprit in my setup. Maybe it’s the same in your setup and we can isolate and help to resolve the issue.
My network topology
(ISP) --- [router] --- [2.5GbE hub] +--[MT6000 dumb AP-1] -- (lan port) -- [linux server]
|
|
+--[MT6000 dumb AP-2]
MT6000 Openwrt (OpenWrt SNAPSHOT r26043-1991bfb814, Version before packet steering was changed]
WPA3 & 802.11r disable (802.11r also had various problems (known issues with WPA3) and was eventually disabled)
The topology is very simple.
Problems that occurred after moving two APs.
- Servers in the lan under the AP cannot be accessed.
- Extremely slow access to sites on the internet.
(Not all sites are slow).
Particularly noticeable on (my wife's) iPhone 15 Pro. I upset my wife B)
Restart the bridger on all APs when this occurs. It will then return to normal.
WED and bridger are now disabled.
1 Like
Similar situation for @apccv and myself. Thank you so much for explaining your setup. I have the same issue, and restarting the dumb AP solves the issue, but as soon as I start moving from dumb ap to dumb ap, the issue pops again. I hope we can help and get this issue resolved.
@gssjshark, @nxhack, and @apccv -- I wonder if you might consider starting a new thread specific to these issues you're seeing. It might help get more, and better focused, visibility into your issue--perhaps gaining suggestions even from other (read: non-MT6000) device owners who might have ideas.
Regardless, I find this very interesting because I have 3 x MT6000s and in my current build (r26086-1b190dfd3a) I run 3 x 2.4Ghz SSIDs, 2 x 5Ghz SSIDs, WPA3 enabled on two of the three SSIDs, WED + Bridger, 802.11r/k/v with Usteer, etc. and don't see the issues you guys are mentioning. Not at all claiming you're mistaken about seeing the issues you see, but I suspect there is something bigger in play here. Again, perhaps a new thread would be a more ideal place to dive into those specifics and nuances.
I am especially interested in seeing if we can figure this one out. I also seem to be unable to effectively communicate with hosts that are connected directly to my MT6000s. (e.g. See Here) I have done some testing/experimenting with various hairpinning settings, to no avail. Would be great to get this one figured out.
On a related, but somewhat hypothetical note... I am debating making my build available, sans some things like SSH keys and such that I bake in, for testing. Not sure if that would help or hurt the troubleshooting. Just thinking out-loud right now.
2 Likes
Oh, guys and gals...
Curious if you've seen this and if it might be leading to the issues you are
experiencing:
TL;DR: Once I removed option multicast_to_unicast_all '1' from all my SSIDs, many of my client connectivity issues stopped.
GL issue opened for this here:
4 Likes
Great deal on GL-MT6000 at Amazon UK. Now £119.92!
Lowest price so far. Last deal was £127.
3 Likes