We are not 100% sure this is related but our server was recently attacked and to solve the problem we locked down all inbound ports.
A few days after the attack we noticed some of our processes would not launch which is triggered by an internal NodeRed flow using IP 127.0.0.1 to access the local Camunda API.
In a test we opened the inbound ports and we got a different error which gives us the idea some inbound port is needed for Camunda processes to run, even though we are accessing internally using 127.0.0.1
Often if there are 2 posts asking the same question it can be harder keep the context of the discussion and the eventual answer in one place. I closed the other thread because when you get an answer it’s more helpful for people in the future to find and understand it if it’s all in one thread.
Sounds good… I idea was to treat both post like documentation and inter-link them for future searchers. If you noticed the people were already posting on the last post looking for a solution.