2.9.10 Problems (RB532 + CM9)
Posted: Tue Jan 10, 2006 7:37 pm
We recently upgraded all our Mikrotik boxes to 2.9.10, all of which were running 2.9.6 prior to the upgrade. Some problems have been introduced with the update that is causing major headaches for me and my customers.
On nearly all of our Mikrotik boxes we are now having disconnect issues both on the backhaul links and on many of the local clients that connect.
The backhaul connection is CM9 + 24dBi Grid to CM9 + 24dBi Grid, signal strength is -65, and the connection was running for 4 months without issues until recently. The backhaul connection has been dropping periodically thoughout the day, the log shows the following:
01.02.03A: lost connection to 'MAC', no beacons
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , got deauth: class 2 frame received (6)
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , assoc failed: unspecified (1)
01.02.03A: connected to 'MAC' on 2437, SSID 01.02.01A
This link now will disconnent and reconnect with exactly the same message 5-10 times per day taking anywhere from 40s to a few mintues to reconnect. The 01.02.01A AP has only two connections to it, this one and a second with a similar setup, it too has the same issues but the times they happen do not coincide.
The same box also runs a 2nd CM9 + 12 dBi omni for customers to connect. About half of the people connected to the omni are also having problems periodically thoughout the day, with a few customers disconnecting and reconnecting upwards of 20 itmes per day the following log info shows the different errors I am getting on the client side:
Condition 1
01.02.03B: disconnected , 'MAC1' reassociating
01.02.03B: data from unknown device , 'MAC1' sent deauth
01.02.03B: 00:0B:6B:36:BB:93 connected , 'MAC1' wants WDS
Condition 2 (Typically 2-3 minutes until reconnect)
01.02.03B: disconnected 'MAC2', decided to deauth: group key update timeout (16)
01.02.03B: 'MAC2' connected, wants WDS
01.02.03B: disconnected 00:0B:6B:36:14:C8, reassociating
01.02.03B: data from unknown device 'MAC2', sent deauth
01.02.03B: 'MAC2' connected, wants WDS
Condition 3 (Anywhere from 30s to a few minutes to reconnect)
01.02.03B: disconnected 'MAC3', extensive data loss
01.02.03B: 'MAC3' connected, wants WDS
Condition 4 has the same 'no beacon' error as in the backhaul connection
All clients are using RB532 + CM9 + 14dBi Rootenna's
Clients signal strengths range from -60's to -84
All AP's are running in 'ap bridge' mode, and all clients are using 'station WDS' to connect and we are using PPPoE for user authentication.
Has anyone else experienced any of these issues?
On nearly all of our Mikrotik boxes we are now having disconnect issues both on the backhaul links and on many of the local clients that connect.
The backhaul connection is CM9 + 24dBi Grid to CM9 + 24dBi Grid, signal strength is -65, and the connection was running for 4 months without issues until recently. The backhaul connection has been dropping periodically thoughout the day, the log shows the following:
01.02.03A: lost connection to 'MAC', no beacons
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , got deauth: class 2 frame received (6)
01.02.03A: failed to connect to 'MAC' , join timeout
01.02.03A: failed to connect to 'MAC' , assoc failed: unspecified (1)
01.02.03A: connected to 'MAC' on 2437, SSID 01.02.01A
This link now will disconnent and reconnect with exactly the same message 5-10 times per day taking anywhere from 40s to a few mintues to reconnect. The 01.02.01A AP has only two connections to it, this one and a second with a similar setup, it too has the same issues but the times they happen do not coincide.
The same box also runs a 2nd CM9 + 12 dBi omni for customers to connect. About half of the people connected to the omni are also having problems periodically thoughout the day, with a few customers disconnecting and reconnecting upwards of 20 itmes per day the following log info shows the different errors I am getting on the client side:
Condition 1
01.02.03B: disconnected , 'MAC1' reassociating
01.02.03B: data from unknown device , 'MAC1' sent deauth
01.02.03B: 00:0B:6B:36:BB:93 connected , 'MAC1' wants WDS
Condition 2 (Typically 2-3 minutes until reconnect)
01.02.03B: disconnected 'MAC2', decided to deauth: group key update timeout (16)
01.02.03B: 'MAC2' connected, wants WDS
01.02.03B: disconnected 00:0B:6B:36:14:C8, reassociating
01.02.03B: data from unknown device 'MAC2', sent deauth
01.02.03B: 'MAC2' connected, wants WDS
Condition 3 (Anywhere from 30s to a few minutes to reconnect)
01.02.03B: disconnected 'MAC3', extensive data loss
01.02.03B: 'MAC3' connected, wants WDS
Condition 4 has the same 'no beacon' error as in the backhaul connection
All clients are using RB532 + CM9 + 14dBi Rootenna's
Clients signal strengths range from -60's to -84
All AP's are running in 'ap bridge' mode, and all clients are using 'station WDS' to connect and we are using PPPoE for user authentication.
Has anyone else experienced any of these issues?