Was looking at contributing on this issue, and looked through the PR referenced. It looks like those were debug logs added when some kind of default variable or value was being provided, or a special action was being taken due to arguments into the task. The docker tasks don't seem to follow this same pattern. Thoughts on where to put additional logging?
Ah yeah @alexisprince1994 this issue isn't very descriptive. On each of the Docker tasks they use the Docker low-level API and each function generally comes with a flag for some type of output (e.g. default to stream=False). Each of these Docker API calls should have some better logging attached!
Example: in the PullImage task the end of the function returns the string output from pull but with stream=True we could watch the output, live log it using the prefect logger, and then once the generator is complete return the output.
Another example in the BuildImage task we capture the output and return it but some better logging could be added to notify users of the Task that the build is actually happening .
Another another example (not relating to images but to containers) something like the StartContainer task doesn't have any Docker API output that the function returns but an extra log such as Docker container started with ID ...container_id... would be really useful!
Yea I vote we log a good portion of the output streams from the various Docker API calls, specifically with a focus on debug logs that help users see how / when things go wrong.
La fusi贸n exitosa de una solicitud de extracci贸n puede cerrar este problema.
I will look into this issue