Ok, I know I'm a broken record but I strongly believe that more noise you make the more people hear you... yesterday I noticed my one of my WiFi APs was dead - PoE device was not working and cause a rift of Google Home failures. I would have loved to know this preemptively by having a notification sent or a device status tied to it via ping response. ICMP PING PLEASE.... I BEG YOU ALL ! BTW I do not want to use the Web Pinger it never works right nor is it lightweight. Also it does not have to be complex just work once every 5 secs or more. Ping is the most basic diagnostic indicator in the Network world, no analytic automation engine should be without it.
You should find the code you need in this driver, just copy or use in a Rule.
2 Likes
I got all excited thinking they opened up the ping calls. This just does an asynchttpget. Similar to other ways. We are still stuck without a true ping.
I would love to see an arpping. I use that in a docker to monitor network presence of devices. Even works if my iphones go into sleep mode and the response to it being on/off the network is almost instant. Works so much better than ping in my case.
The get is to send the command to the hub, the hub then executes the ICMP ping, and returns the results...
2 Likes
Have you considered a simple Network Monitoring System? Something designed for the task?
3 Likes
Why, everything I need is in a simple analytical box like Hubitat.... NMS require servers or appliances - quite a heavy source for home automation. Servers and appliances fail and they don't run under power loss\UPS well. Hubitat is sips the power.
I generally agree with this philosophy. However, for some devices I test whether they are up or not before cutting power to them. Ping comes in handy for that.
1 Like
I can't agree with this. If by "server" you mean some form of sizable box, an NMS definitely doesn't require such. And the Hubitat is an appliance. A Pi or a Beagle appliance is just fine for an NMS. These units sip power at a similar level to a Hubitat, and you will get a lot more monitoring value out of it than just having ICMP on Hubitat.
I run my NMS on a Big and Powerful Atom chip. 
2 Likes
Certainly not arguing against general use of ICMP. How could I? I'm responsible for trillions of ICMP packets around the world every day... 
I am however arguing against asking the Hubitat folk to implement a driver for it so people use Hubitat as a ping monitor. So many other things that they could be working on closer to Hubitat's core value.
5 Likes
Agreed. And I actually don't do my pings using Hubitat; just use the outcome to control Hubitat-paired outlets.
2 Likes
This does not make any sense as ping is a true indicator of the network. There are very many who don't want to or have the ability to run NMS. NMS is saying I going to have a car and be electric only, while others use gas. Why would you have two boxes that one can handle a same task. Simply if the home world is going for a single control engine of house devices to manage then ICMP pinging devices does make sense to build internally for automation. Development would be worth it.
ooooooo... 1 step closer. I see now. Sucks that it takes an http call to the hub for the hub to do it. But interesting... can use this for some stuff I was looking to do. Thank you.
1 Like
Because each is designed to a task, and the tasks are rather disparate. You got here because of a PoE failure in your networking infrastructure. Monitoring your network physicals, switches, access points and the like is a good use case for an NMS. It's not a good use case for a Z-Wave/Zigbee automation bridge.
It's possible that the Hubitat folk will do it (ICMP) some day... but I wouldn't hold my breath. You might consider moving on to another solution.
3 Likes
It would require hardware capable of handling a lot of events per unit time. Much much more than generated by zwave and zigbee sensors. IMO, NMS is a use case for different hardware.
2 Likes
What network management software do you guys like to run?
I've run a few. Settled on Librenms a few years ago. Good SNMP support for hosts, switches, access points, and the like. Built-in alerting. Docker hub distribution.
1 Like
Opinion - Not a supporter of the idea. The world is full of products that try to solve too many problems with one answer. Like nails, not all hammers are right for the job.
While recent addition of some basic IP network stuff has been added, it's an attempt to be pertinent to the needs of the HE. I'd rather see a more robust file management feature with cifs or something!
As to the question of 'which NMS'. a very old product called Dude was written years back by a router OS company and it's still my favorite solution. Logging, support of port monitoring, and SNMP. Visual display of net with drill down capability. It's old. but it's free! (i've images of a 320 camera network with a 10G backbone I built for a state prison but I can't find them!)
In today's world you must Elon Musk a product. What sets it apart from the rest. 2nd Ping isn't much of a ask to get working. No reason that this device can't have it built into a rule or a device indicator. Not everyone will run a NMS and also these types of boxes can be placed in a business application where other servers are not deployed and reduce costs.
1 Like
Well this is a 7 month old post, they have added icmp calls since then, I think it’s even an action in rule machine now.
4 Likes
Considering this thread is many months old.... they added the routines for developers to utilize ping in apps/drivers. And they added a ping action to rule macine. Hopefully thats what you are looking for.
2 Likes
