AP541N restarting when loaded

[trackhappy]

Hi All,

We have 9 AP541n's in a cluster. There are two in an area where there is a high concentration of devices (high being potentially 50-ish on an intermittent basis when Parliament sits). We believe there is only about devices 10 trying to connect at present and can only see 6 connections in the console after the recent reboot. These two units have been running along without incident until this week when the clients started a session (previously only 1-2 clients present). It seems that first one unit crashed/rebooted, then the other one in the same area rebooted soon after. I am guessing the second one rebooted when the clients all tried to connect to it after the first one crashed. Others in the cluster have remained operational and unaffected. They are running 2.0 code. We have been through the "DHCP issue" and this new versions seems to have cured that one, but the crashing under load is new.

They are pretty much default except for IP addresses obviously, a single vlan only, and using Radius mac authentication to a Microsoft RRAS server.

We logged a call with the Cisco helpdesk over other issues along the way (date and time resets, radius entries don't appear in the log until after an AP reboot) and got some person on the phone who eventually decided that they don't support Microsoft Radius, despite us trying to explain to him that it is nothing to do with the issue. He had no idea, no skills and didn't want to help. I am somewhat less impressed by Cisco now than I was. So....

Anyone got any ideas?

[trackhappy]

As Leon said, we still have the problem. Most of the 10 AP's reboot randomly with no obvious pattern except perhaps a period of inactivity then a little load and off they go. It may be iPads connecting that cause it but we can't be sure. Haven't loaded 2.02 but somebody else here said they have.

We are looking at replacing the whole infrastructure here shortly anyway so have basically given up trying to get it fixed.

Suffice it to say that the wireless infrastructure going forward is unlikely to be Cisco given our experience. And before anybody says it, I know they are not "real" Cisco, but they have a Cisco badge on them and that was one reason I chose them.

[halcanites]

We are also experiencing this problem.

We have 10 APs and one of them has been rebooting by itself during the middle of the night on a daily basis, as reported by our monitoring tool.  It is running on firmware version AP541N-K9-2.0(2).  One thing weird though is that the system uptime says it has been up for the last 29 days.

[weichert2]

AP541N rebooting.

We have about 150 of these devices deployed with random ones 'slowing down' and eventually are rebooting. We have been working with Cisco on this since Feburary of this year. (4 months) and still no resolution has been found. It appears to be when we use RADIUS/MAC authentication that seems to have some kind of memory leak that causes this issue. I am not the expert on this but have been sending logs on how the CPU spikes up prior to the device rebooting. It's hard to get logs like this since it locks up prior to rebooting. We are still working with them in hopes of finding a resolution for this issue. We only get updates every other week so I don't expect an answer anytime soon.

Show TOP during quite early morning: (192.168.105.3)

Mem: 45088K used, 11492K free, 0K shrd, 9152K buff, 16192K cached

Load average: 4.00 4.00 4.00

  PID USER     STATUS   VSZ  PPID %CPU %MEM COMMAND

  421 root     RW      1256        144       99.4 2.2 hostapd

  485 root     RW      1452        483        0.1  2.5 top

  380 root     SW      9244        144        0.0 16.2 snmpd

  144 root     SW      5792          1         0.0 10.1 dman

[Kenneth Lee]

We've had the same issue.  It seems that disabling SNMP has helped resolve the issue.