Ping watchdog that power off AP until uplink restored.

[scracha]

In our scenario we often have solar towers with say AF5x and a redundant backup link(s) such as Airmax M via OSPF. To save power the redundant backup links are switched off until they're required (simple ping watch and on some devices like tycon we can set the AF5x to turn off when voltage is low then the Airmax M will kick in with ping watch).


Only way I can think to do this with Netonix at the moment is script some SSH commands from the attached Mikrotik. Not ideal.

Is there any example SSH scripts or documentation for CLI SIRHC?

EDIT:


So I wasted a few hours on this today. Turns out Mikrotik SSH client support is lousy from a script
https://forum.mikrotik.com/viewtopic.php?t=71262
Should be easy enough from a linux box with except script but doesn't help me.

/system ssh NETONIX_IP
configure
interface port [PORT NUMBER]
poe 24V {or poe off}
end
Then expect an enter prompt.

I could shove a mikrotik up there with POE out but that kinda defeats the whole purpose of the netonix. So we can use Netonix on our little solar sites and also our powered sites but at present this is preventing us from using Netonix switches on our medium and large solar sites.

[Eric Stern]

You can already do this. When creating the watchdog set the Action to "Disable power", the specified port will be powered off as long as the ping is failing.

[scracha]

Ah brilliant. I'll test that out on next deployment!!

[bpeach]

Any news with this?

This is not possible with the current watch dog rules, to my knowledge anyway.

This would require a rework of the watchdog feature which is something that has been thrown around - Making a more advanced type of watchdog similar to a script or advanced logic triggers.

This might be something we will do in the future but is not on the immediate list of things to do.
...

Sorry to raise an older topic, but I have a request that is somewhat related to this... and I don't see it in any of the older threads.

I use (and LOVE!) the WISP switch series products at my wisp. Typically, I use the DC models (even at A/C powered sites) with a pair of AGM batteries and a Tycon UPS/charger.

I would like to be able to monitor A/C power at my various tower sites... The existing "watchdog" feature on the Netonix is 99% of what I need since it already allows me to generate email alerts without taking any further action... unfortunately it only allows me to trigger alerts based on ping failure..

Using this simple (and cheap!) trick (refer to http://gregsowell.com/?p=2093) I can generate an ethernet "link down" when the utility power goes out. I would like to be able to make use of available ports on my Netonix switches to trigger an email notification based on eth link down using the watchdog.

It would also be really cool if the email could include the Description field associated with the port/device that is being monitored. Currently it only gives the IP address that failed ping. This would be extremely handy, even when using watchdog for ping monitoring.

Thanks in advance!!

Ben
Zetalink

[Eric Stern]

You really should be using an NMS for this sort of alerting, its what they are for. The watchdog feature was not designed for this sort of functionality, it was intended to power cycle buggy radios in case the link went down.

LibreNMS is quite popular, free, and supports Netonix switches out of the box.

[sirhc]

It would also be really cool if the email could include the Description field associated with the port/device that is being monitored. Currently it only gives the IP address that failed ping. This would be extremely handy, even when using watchdog for ping monitoring.

So your asking that the email alert contain the Watchdog rule description?

If so that I am sure Eric could do easily and makes good sense.

[bpeach]

So your asking that the email alert contain the Watchdog rule description?

If so that I am sure Eric could do easily and makes good sense.

Yep, exactly... This would be very helpful for us.

Currently the body of the email looks like this:
Failure checking 10.64.0.38 - from GrassyKnoll Netonix (10.4.250.2)

I think it would be much nicer if it looked like this:
Failure checking 10.64.0.38 (AF24 link to HamTower) - from GrassyKnoll Netonix (10.4.250.2)

Obviously the text I added in the example above comes from the "Description" field in the Watchdog rule.

Please and thanks!

Ben

[bpeach]

You really should be using an NMS for this sort of alerting, its what they are for. The watchdog feature was not designed for this sort of functionality, it was intended to power cycle buggy radios in case the link went down.

LibreNMS is quite popular, free, and supports Netonix switches out of the box.

OK understood. Valid point too.

It was just an idea.... I know a lot of folks are using Mikrotik devices with a script that sends an email alert on ethernet link loss... I just thought it would be a cleaner solution if this was built in to the Netonix. I am certainly not arguing your point about using NMS; I am simply telling you what some folks are doing today.

To reiterate the "big picture" here, I think it would be a great selling point for the Netonix WISP switch series (especially the DC models) to support some sort of A/C power monitoring/alerting. My idea of doing this by monitoring an ethernet link is just one way of achieving this. Another way might be to utilize the serial port. All I'm doing here is trying to leverage your current hardware to add a valuable new feature that could help WISPs (and ultimately drive switch sales).

To me, A/C power monitoring is a pretty basic function (every WISP wants to know when the power is out at a tower site) and it seems like something that could be supported on your platform with relatively little effort since you already have a watchdog with email alerting.

To your point, I suppose one could monitor for Ethernet link loss via an NMS application and trigger alerts accordingly....

Thanks anyway,

Ben

[scracha]

You can already do this. When creating the watchdog set the Action to "Disable power", the specified port will be powered off as long as the ping is failing.

Hi Eric,

Finally I've got a new solar powered tower needing redundant backhaul that I only want activated when the main link goes down.

Apologies if I'm missing something. The "disable power" method described above is doing the exact opposite of what I want.

I want to "Enable Power" when a ping is failing.
AND
I want to "Disable Power" when a ping succeeds.

[sirhc]

I asked Eric to respond to your question