I've had an Airrouter HP running for a week now, with the beta firmware. It is configured as a gateway and as a tunnel server, with 10 tunnels enabled, 3 of which are active. There are about 40 nodes connected to this one, mostly over one of the tunnels. I'm using the UBNT PoE supply.
Over the course of a couple of days, accessing the web interface becomes progressively slower. Last night, I couldn't navigate between pages. Eventually the tunnel I was on shut down, and while trying to SSH into the unit directly using a locally connected computer, it showed a 'bad gateway' message in very large text on the web interface, and then rebooted itself. After the reboot, it appeared to operate normally. Normally, when it slows down like that, I reboot it myself. This is the first time it's reboot on it's own. During these episodes where it's noticeably slow it also seems to be very slow at passing data to and from other nodes it's connected to. During this latest event I never got a chance to see the status page to see what the memory stats were. I'll watch that from now on.
I haven't noticed any other physical issues, such as high heat that could explain this kind of problem.
Cheers,
VE3ZX
Over the course of a couple of days, accessing the web interface becomes progressively slower. Last night, I couldn't navigate between pages. Eventually the tunnel I was on shut down, and while trying to SSH into the unit directly using a locally connected computer, it showed a 'bad gateway' message in very large text on the web interface, and then rebooted itself. After the reboot, it appeared to operate normally. Normally, when it slows down like that, I reboot it myself. This is the first time it's reboot on it's own. During these episodes where it's noticeably slow it also seems to be very slow at passing data to and from other nodes it's connected to. During this latest event I never got a chance to see the status page to see what the memory stats were. I'll watch that from now on.
I haven't noticed any other physical issues, such as high heat that could explain this kind of problem.
Cheers,
VE3ZX
Do you hae any Linksys devices on your system? They seem to introduce some instability as well. I'm trying to get all of them off of our system.. at least where they are performing any critical functions like Tunnel Servers and RF gateways. Probably ok as an end user device for connecting in, but that may be questionable too.. I imagine that the "days are numbered" with regard to the old WRT54G units. Since the AirRouter AR at around $36 and the HP around $57, it's pretty much a "no brainer" to finally phase out the old Linksys devices.
I haven't evaluated the thermal operations of these units, but I will pay closer attention to that.
I'm eagerly awaiting Beta 2, but wanted to report this in case others are seeing similar behaviour. The UBNT stuff does seem to run warm, but the AR isn't any warmer than a pico or bullet.
73
When the poor response conditions begin to occur, try to capture the support data, say every 6 to 8 hours as it degenerates--link at bottom of Administration page. It's optional to be up at 3AM to do a capture :) . Let's try to capture a trend before it becomes unresponsive.
A possible related symptom is work to address lack of sufficient memory during "keep settings" upgrade scenarios. We are addressing this (with time-to-release in mind) by shutting down services/resources before upgrade to have confidence this will work. Let's hope this situation is something other than memory running wild issue. If this is memory related, or even more memory related symptoms show up, then we'll be motivated to get to the bottom of it now and extend the beta period...
Let's start by capturing the support data. If more information is needed, we might try connecting a tunnel between us for direct access.
Joe AE6XE