https://dagster.io/ logo
z

Zuhair Ikram

03/05/2021, 1:49 AM
I was wondering if this sort of a set up is possible with
dagster
? The other two boxes represent hosts with like an agent or a “lite” version of dagster. The agent kicks off processes on the host when main dagster scheduler tells it to.
n

Noah K

03/05/2021, 4:39 AM
As in you want to implement solids in both arbitrary python and dbt?
z

Zuhair Ikram

03/05/2021, 6:00 AM
Yes
n

Noah K

03/05/2021, 6:17 AM
Sure, solids can be written however you need 🙂
z

Zuhair Ikram

03/05/2021, 6:34 AM
Sorry, let me rephrase. Would I basically do a
pip install dagster
on all three boxes +
pip install dagit
on one box to get this going?
Is there a write up on how one would set up what I have diagrammed there?
n

Noah K

03/05/2021, 6:37 AM
running things on multiple servers is unrelated to the language you are writing your solids in
Look at things like dagster_celery and dagster_k8s for multi-node execution options at varying levels of complexity 🙂
z

Zuhair Ikram

03/05/2021, 8:07 AM
Yeah, I think we went down the wrong path for my question. It wasn’t about the language for solids, but more about a multi-node installation. I was trying to compare the complexity for implementing a multi-node installation for dagster vs some alternatives like Airflow2, etc.
n

Noah K

03/05/2021, 8:10 AM
Overall fairly similar
Celery on Kubernetes is the usual way forward 🙂
You can mix-n-match it a bit from there
z

Zuhair Ikram

03/05/2021, 8:16 AM
Thanks!
n

Noah K

03/05/2021, 8:17 AM
(Dask is also supported but I hateses it and will literally never recommend it 🙂