I was thinking about purchasing a FlowForge subscription but I couldn't understand if the HA mode meets my expectations. What happens when an instance fails/dies? How do they achieve Load Balancing?
For example if I have a chain with 11 nodes and my instance dies during the fifth one, I want it to continue from the fifth one(not from the start) when I run it again. Will they offer me that? Or is there an open source solution?
FlowForge HA is about highly available runtime. I.e. there are multiple node-reds running the same instance.
This provides both HA and horizontal scaling.
See the release info here: FlowForge now offers High Availability Node-RED • FlowForge
This is more like Highly Available Nodes. Every node would have to be re-designed with high availability in mind. This is not likely to happen!
So, in essence, this is really not something that can be easily solved with a simple "option switch". You would need to engineer the feature in your flows. But you can achieve this with the HA in flowforge.
For example, instead of 11 nodes, you would use 11 "project links" and those would route to a subflow/subroutine on the available instance. Each subroutine would then be ran on the available instance. If an instance falls over part way through, the HA would route the messages to the other instance.
Even then, there is no guarantee the type of failure you are trying to guard against is the failure that occurs (ie this would NOT be 100% fail proof).
A better solution is to use a queue and "peek" the msg. If it doesn't reach the end of your flow, peek it again until it is fully processed - then pop it from the queue & discard it.
@Light2K95 if you head over to the FlowForge forum, you will get more relevant support. There, we can pin point your exact requirements and see if we can fit your needs. I am sure we can help you achieve a level of HA that works for your requirements.
Please link back to the new topic for others to follow