https://dagster.io/ logo
Title
s

Sterling Paramore

08/30/2022, 10:34 PM
Started getting this error today during a serverless deploy. Not sure how to interpret it. Is it something wrong with ECS?
The gitlab CI script is running this when it fails:
dagster-cloud workspace add-location --organization=${DAGSTER_CLOUD_ORGANIZATION} --api-token=${DAGSTER_CLOUD_API_TOKEN} --deployment=${DEPLOYMENT_ID} --package-name=eta_wh --image=${REGISTRY_URL}:${CI_COMMIT_REF_NAME} mainspring_warehouse
This command has worked before. This was all working this morning. I’ve been making some changes during the day, but nothing that seems related to this kind of error.
@prha - Any chance something you did to make the base image customization work for my org that would be causing this now? I did get that base image customization working yesterday.
p

prha

08/30/2022, 10:45 PM
hmm, nothing changed on the serverless image building side that I know of… lemme do some digging.
Is this a persistent issue, or was it ephemeral? I’m checking the logs and the user code launcher seems to be spun up
s

Sterling Paramore

08/30/2022, 11:18 PM
Welp, I re-ran an old ci pipeline and it seemed to work, so I guess I did break something. I’ll try walking through the commits I’ve made to see what went wrong.
Ok, so it looks like this happened when I changed the name of my package from
mse_wh
to
eta_wh
. Seems like that’s all I did and I don’t know why that wouldn’t work
Ah ha! It wasn’t that either….. I also changed the
ENTRYPOINT
from
[]
to
["/bin/bash", "-c"]
. When I use
[]
, it works. I needed the other one for working with the container locally
:daggy-success: 1
💡 1