Flower (Celery) Dies on Fresh Project

After configuring my cookiecutter project and running docker-compose up -d as specified, everything starts EXCEPT for the flower service, which prints the following message:

ERROR: for flower  Cannot start service flower: OCI runtime create failed: container_linux.go:370: starting container process caused: exec: "--broker=amqp://guest@queue:5672//": stat --broker=amqp://guest@queue:5672//: no such file or directory: unknown

Cursory internet searches make me think that this is using the wrong broker address? Looking at my rabbitmq server (which - by the way - I had no idea I had to install and only inferred from the above internet search) logs, it appears that my broker is running at rabbit@localhost:

*excerpt*
2021-07-21 16:38:28.986 [info] <0.274.0> Running boot step cluster_name defined by app rabbit
2021-07-21 16:38:28.986 [info] <0.274.0> Running boot step direct_client defined by app rabbit
2021-07-21 16:38:28.986 [info] <0.274.0> Running boot step rabbit_management_load_definitions defined by app rabbitmq_management
2021-07-21 16:38:28.987 [info] <0.548.0> Resetting node maintenance status
2021-07-21 16:38:28.987 [info] <0.44.0> Application rabbit started on node rabbit@localhost
2021-07-21 16:38:29.034 [info] <0.44.0> Application rabbitmq_management_agent started on node rabbit@localhost
*more logs...*

However, changing occurrences of guest@queue to rabbit@localhost doesn’t fix the error.

Side note: I have no prior experience working with Celery, and am honestly confused as to why it’s included in this stack - would love to be enlightened.

[markkvdb]

This problem is related to the v1.0 release of Flower. I’m not quite sure what Flower changed but it’s based to set the image version explicitly to a version that’s working. In my case, I set the flower image to flower:0.9.7 since this is the last version that still works for me.

[waclock]

Fixing version to 0.9.7 worked for me as well.