I created a rule action to send a http-get-string. This works fine to reach a host within the LAN (per IP address).
But the get-string for a remote server (per domain name) doesn't work. The hubitat log shows, that the action is processed without errors. Nevertheless, the target host isn't reached (no entry in the access log).
The used URL is working from a PC with Firefox. The LAN is a windows domain network, where a local dns server is running. The firewall of this server is open for all ports of hubitat. The Gateway router is a fritzbox without an outgoing firewall.
So my only explanation is, that hubitat doesn't send out the get-command to a remote Host or to a domain?
Is there another way to send the get command? Is there a setting to switch?
That might be your explanation; however, others and I have several built-in and community integrations that send http get commands outside my LAN, and work fine.
So the explanation lies in some unique set of circumstances relevant to your hub in your LAN.
Helpful and reassuring to hear that remote commands work in other installations!
So I have to find the barrier in my environment. Thought at the moment I miss a real starting point.
I would test in a test driver (simple one command and two methods (sendCommand, parse). The parse method will provide more information, such as response.status and response.data) to see what the return actually contains. (PS: review the documentation for parameters that may help.)
You may also get more information of the response using your browser and the view source function return. It may not be what you see on the main page.
I started with hubitat 3 days ago. So driver programming will be a challenge - and a chance.
But I fear the problem is, that the command is absorbed anywhere in my LAN (without response). The remote server has no entry in the access log.
Something like tracert would be helpfull.
This Traceroute Test shows no problems. And important: The only local hop is my Gateway Router. I expect this is not different for the get action.
So the problem must be located in the gateway.
Tracert from the hub will be the same route a request from RM should take, my guess is that you may not have the Get setup correctly in the rule.
I figured out that a surprising bug caused of my http-get problem:
Setting a fixed IP for caused the DNS resolution of the hub to fail (independent of DNS-server setting). Switching to DHCP immediately fixed the problem.