Is it normal for a brand new airflow installation to take multiple seconds to run a simple DAG? #800
Unanswered
albertjmr
asked this question in
Questions & Answers
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I recently deployed airflow to a company managed k8s cluster using this chart!
Everything is running fine and I can already run some dummy DAGs, etc. I'm mostly using the default values that come up with the community chart (using
CeleryExecutor
).However, I've noticed that even a simple DAG takes many seconds to execute, tasks take always 1-2s in the queuing step for instance.
Here is the gantt chart for a simple DAG I created based on some examples online, as you can see, it takes over 1s to queue each task and the task themselves take seconds to execute too (even though they only involve 1-2 lines of python):
The code of this DAG is as follows:
I know that airflow isn't meant to be extremely quick but I was surprised to see this when there is literally no load in the system yet.
Is this normal? If not, could you give me some pointers into what I can look into in order to troubleshoot this?
Beta Was this translation helpful? Give feedback.
All reactions