After partially localizing my Vera, in an effort to stabilize and only permit proxied access for the iPhone app to work in conjunction with Amazon Alexa, I had noticed some pings still occurring. After digging around a bit, this isnt any function of the NetworkMonitor that Vera uses… but why does the Reactor app do this?
Extracted L_Reactor.lua and discovered a function ‘checkInternetState’ with the following partial config:
local function checkInternetState( pdev )
"8.8.8.8,8.8.4.4,1.1.1.1,www.facebook.com,www.amazon.com,www.google.com"
So suppose I have a few questions here…
What is the purpose of this health check? I have my firewall blocking arbitrary pings.
If there is a genuine purpose, can there be an option to instead ping the default GW IP address? Perhaps this should be a configurable option?
Yup, that answers everything. Dont like using DNS for monitors personally, so hard coded the 1st 3 check IPs and narrowed down the firewall policy. Thanks!
This Network management feature works very well, but I would like to add a suggestion to a feature improvement:
The time not in minutes, but in seconds, as we have in almost all other Reactor functionalities;
The time for the other two tests can also be configured.
I would like to be able to test every 15 seconds an eventual failure, then 10 seconds for the second validation, and then 5 seconds for the third and definitive confirmation that there is a communication failure.
This way in 30 seconds we could already take an action such as changing to a redundancy of the internet connection. In the current days, staying around 3 minutes, which is the minimum we have in the current configuration, can be a long time.