There is serious bridge problem. Mikrotik have messed something in bridging since about 2.9.10 version.
Today I have dowgrade from 2.9.26 to 2.9.8 one of my key AP to stop disconecting all clients.
There is serious bridge problem. Mikrotik have messed something in bridging since about 2.9.10 version.
Today I have dowgrade from 2.9.26 to 2.9.8 one of my key AP to stop disconecting all clients.
Could you be more specific and explain what problem you are seeing, your config, …?
Best regards,
Christian Meis
What exactly is the problem?
When /in which situation/ it appears?
Everything works OK on all my 2.9.26 bridged WDS AP’s
Until yesterday when all 2.4 clients on one AP was disconecting about twice A SECOND on all interfaces. This is PC based, 5 wlans, all bridged.
Logs are totally flodded with something like this:
00:30:…@wlan1: connected
00:30:…@wlan1: disconnected, decided to deauth: authentication not valid
(2)
Disabling and reenabling problem interface from port solves the problem for a few minutes.
Downgrading it to 2.9.8 solves problem permanently.
What does lead you to the suspicion that this is caused by the BRIDGE code?
Best regards,
Christian Meis
00:30:…@wlan1: disconnected, decided to deauth: authentication not valid
Perhaps you need to stop worry a bit about the bridge, read the error message, and then look at your MAC authentication, or whatever authentication you are using that is not allowing the client to connect ![]()
Just a thought.
in next version we will have workaround for this problem.
Uldis,
can you describe what the problem actually is, please?
Best regards,
Christian Meis
Problem is in bridge, because disabling and enabling PORT for bridge for problem interface stops disconecting.
In 2.9.10 there was some big changes in bridging.
could we get some more deeper details ?, cant make anything from this
I’m not a programist and have no time to debug problems. Last week I had to downgrade another machine from 2.9.26 to 2.9.8. I’m glad I keep all versions since 2.8.28
Both machines are PCs. One Celeron 2300 have 5 wlan interfaces Ath 5211, 5213, 1 ethernet, all bridged, Routerboard 14.
Other is Celeron 2100 on Intel 865 GBFL board, Routerboard 18, 7 wlan interfaces, 1 ethernet, all bridged. Atheros 5211, 5213, 5413.
On xx other there is no such problems and I’m not using Routerboard 14/18 on any other machine. Maybe this is Routerboard PCI>miniPCI adapter related problem?
pleace check if with other wireless devices you have same behavior.
i had allmost same expierence with Fujitsu-siemens Amilo M3438G - endless reauth-deauth
set in router logging config - wireless,debug and report back what you see
I’m facing the same problem from 2 months with the bridge between to Wlan interfaces.
kapusta_kiszona have you tried to upgrate to 2.29?
Thank you in advance
Max
Have you tried removing some of the radio cards and testing again? If you are using the high power atheros cards (like SR2 or NMP8602) you have to make special provisions for using more than two at a time. Other cards have similar limitations. It is because the cards will draw more power than the pci bus has. I noticed though that the problem only presents itself when traffic load increases.
It could be that disabling and reenabling your bridge port is stopping traffic, and therefor allowing clients to stay connected. And after you re-enable the port and traffic starts to pass it causes the drop again.
I had exact same problem a while back.
Thank you for the answer.
I have tried to disable cards also port.
I have tried with only two card too but the same issue
I have read in some post that they have solved changing from bridge to router mode. I have planned this trial, I hope that it will solve the issue.
My HW equi is RB532 with 4 CM9 but 2 are disabled and two enable.
Client remain connected but they have sporadic high pings.
If I create a Virtual AP and attach it to a single wireless board it work great but I cannot sectorize the area.