Ping watchdog that power off AP until uplink restored.

[CuninganReset]

Hello.

I am just starting with my very first 6 WISP switches, model 8-150-DC

My setup is a tower with two AP (Rocket M) and one wireless uplink (Nanobridge), today if I lost the uplink the two AP remain up and my clients (mobile devices) connect to it and lost the connection to the server (HTML 5 webpage that reload every couple seconds) .

What I want is to shutdown AP until uplink is restored, with this I can avoid the clients to roam to the "unconnected" AP and they will roam to other tower.

How can I achieve that with Ping Watchdog?

[sirhc]

This is not really achievable with watchdog.

Watchdog is designed to power "cycle" a port when an IP or URL can not be reached.

What you want is to have a specified port POE turned OFF when an IP/URL can not be reached and turned back on when it can be reached.

This could be added in a later firmware but this is not currently an option.

[CuninganReset]

Thanks, will you add it in future?
It is any way I can make scripts on the switch?

[Eric Stern]

Actually, I think you can do it with the ping watchdog as it is if you set "Interval" to 30 seconds and set "Time" to 60 seconds and "Failures" to 1.

This will cause it to do the check every 30 seconds, and if it fails it will turn the power off for 60 seconds. But then it will do the check again 30 seconds later while the power is still off and if it fails again it will reset the shutdown timer for another 60 seconds. This will continue until the check starts succeeding again, then the power should come on again about 30 seconds later.

[CuninganReset]

Removed.

[CuninganReset]

Actually, I think you can do it with the ping watchdog as it is if you set "Interval" to 30 seconds and set "Time" to 60 seconds and "Failures" to 1.

This will cause it to do the check every 30 seconds, and if it fails it will turn the power off for 60 seconds. But then it will do the check again 30 seconds later while the power is still off and if it fails again it will reset the shutdown timer for another 60 seconds. This will continue until the check starts succeeding again, then the power should come on again about 30 seconds later.

Hello, sorry but it does not work, this night i had a uplink fault and the switch do not did what i expect.

The behaviour is like this.
When first power on it gives X seconds to AP to power on, later it start making ping at rated Y interval and if Z responses are not given it shut down the port for T time.
But after T time, it start the process from begining, i mean, after T power down time it waits X seconds and later start making ping at Y interval.
This mean that the AP is ON for X+Y*Z time and shutdown for T time and later power on for X+Y*Z another time and shutdown for T time.

Is not the same i am asking for.

What i am asking for is:
On system power on, it waits X seconds to give time to uplink to be established.
Later it start making ping at rated seconds Y and if fail Z times it shut down the AP until ping response is OK for at least Z times at Y seconds.
What i did now, i configured the whatdog as this:

With this configuration i ensure that the AP will not radiate SSID and reboot it before any client connect to this AP.

¿Can you develop this tool or do you think is not needed?

[sirhc]

Eric says he is not following you on this CuninganReset

Maybe a little more detailed explanation will help Eric wrap his head around what your trying to do.

Maybe include a screen grab of these Tabs: Status, Ports, Watchdog, Device/Status

[CuninganReset]

Hello, thanks for your answer.

The purpose or what i want is to shutdown APs forever until the uplink is not ok.

For example a switch with one AP in port 1 and a wireless uplink to the wisp network on port 2.

If the wireless uplink on port 2 breaks(data not flowing), the AP will continue accepting clients but the clients can not reach the wisp network because the uplink in port 2 is broken.

The client will not have internet but the wireless connection, client to AP is good.

What i need is to shutdown the AP on port 1 until the connection to the network via uplink in port 2 is restored, once is restored the AP must be powered on.

With this the client will lost the connection to AP and roam to other AP in view, in my case to other tower ready, i have 6 towers in place that cover all my yard.

If i don't manage to explain myself please ask my i will make a diagram of something, my English is really bad.

[Eric Stern]

I understand what you want to do, but I still don't understand why the configuration I suggested ("Interval" < "Time") isn't working for you.

[lligetfa]

Actually, I think you can do it with the ping watchdog as it is if you set "Interval" to 30 seconds and set "Time" to 60 seconds and "Failures" to 1...

Did you test it? I just now tried using the IP of my phone as the watchdog on a Pico. When I turn off the WiFi on the phone, the Pico goes down but does not stay down.