I've buildt a relative large wireless network for an camping resort, and I have some troubles with stability.
This is how the network is buildt:
Total of 5 AIR-AP1231G-E-K9 Cisco Aironet AP1210 802.11g IOS software, 54Mbit 802,11g AP's with 2 AIR-ANT2506 - Cisco Antenna, 2,4GHz 5,2 dBi Omni Dir. Antennas at each AP.
How the network is connected:
www.loxus.org/camping.gif
2 AP's wired at the main bulding, then 3 AP's in "series" wireless (repeater -> parent) covering almost the entire area. (2 more AP's are being planned to cover 100% of the site).
My problem
After 5-6 weeks of uptime, the 1st (main) parent AP seems to start getting buggy. All the AP's using that one as a parent (or "grand parent"?? lol) loses the connection for about 10-20 seconds, at least 50-60 times a day. When the main AP is rebooted, everything seems to be working 100% stable for another 5-6 weeks.
I have experienced the same thing with cheap D-Link and Asus AP's also, what is the cause for this? Is there any settings I can change, or do I have to create a schedule that reboots the AP once a month?
I forgot to mention that my college mentioned something about stacks... logs.. something that builds up over time - If that can give you any hints...
Thanks for all your help in advance.
Updated 20.07.2006:
I will come back to you with the versions numbers... I upgraded them to the latest version avaliable around march 2006 I think. Don't know what that was. I'll have to get to the site to check this out tomorrow or something.
It's IOS Software yes (if you look closer in the first post
The symptoms are so easily that everyone gets disconnected from the radio/network. In WinXP you get that annoying bubble stating that there is no awaliable wireless networks around. After 10-30 seconds you get connectivity again. The radio is nowhere to be found, in winxp wlan utility, nor in netstumbler / swscanner during these 10-30 seconds.
This affects all clients connected to that AP, and every AP that uses that one as a parent (repeater). This same thing is happening on all the AP's, but AP#1 is the one with most traffic (since it holds AP#2 and AP#3's traffic also) so it's happening more frequent there.
To perform an cold restart of the AP('s) let them go for another period of weeks without problems.