Are there any best practices to follow to deploy python worker to production?
Temporal is not intrusive when it comes to your worker process deployment. Typically what I’ve seen is users build the image and deploy it the way you like (on k8s for example).
Important thing is to use Temporals feature, High Availability, in that workflow executions are not tied to specific worker so having at least 2 worker pods up is recommended for that (in case one becomes unavailable execs can finish on another one).
The other thing is resources you give to your worker pods. Depending on your configs like worker cache settings, if you have very cpu-intensive activities for example, and the load these workers have to do, watch your cpu and mem utilization closely to make sure you give worker processes enough resources to do their work.
Also, some of the info in this checklist is relevant to Python:
After they’re deployed, this guide is for tuning/scaling:
Thanks. Will look into those documents.