(@jgarridoa - is this case? or are you suing RMQ + starting a daemon?)
Yes, this is my case. I wanted to have a similar experience to the previous AiiDA versions so I chose to use RMQ as well. I opted to choose sqlite instead of PostgreSQL just to check if it would feel simpler than PostgreSQL without any significant disadvantage. Now that I know this, I guess I will try to move to PostgreSQL.
@jgarridoa if you want to try a quick hack, you could try to just go here in your code, and put a
returnas the first line of theset_process_state_change_timestampfunction, so the function never does anything. Then, try to run again and it would be great to report if you encounter other issues, or if this solves all problems.
If this works, we’ll open an issue and discuss/plan for implementing this in AiiDA.
Let me try. I will post again reporting the result.
Are you aware of the aiida-submission-controller? Even if you now don’t want to migrate your scripts to it, we would be very interested if the submission controller would work for your use case or if you see features missing, as we are planning to officially recommend it when you want to submit max
Xwork chains at any given time, and submit more only when some have finished, keeping the total number<=X.
Yes, I am aware. In fact, I was encouraged to use it in this post.
I decided not to use it because for my use case, it was faster to write my own python script than learning how to use someone else’s work. I will probably migrate my scripts in the future, when I have more time.