Add support for stargz/estargz images

[bdalpe]

Describe the current behavior

Putting this in the backlog for future consideration.

Our current base image is roughly ~275MiB compressed before any additional code is added. For environments like AWS ECS Fargate, this image is pulled in entirety every single time a container starts which adds some latency to startup times.

docker manifest inspect -v prefecthq/prefect:3-latest | jq -r '[.[0].SchemaV2Manifest.layers[].size] | add' | numfmt --to=iec-i

272Mi

Most containers won't require the usage of every single byte in the image leading to wasted network bandwidth.

Describe the proposed behavior

By implementing this, we could speed up container start times and reduce the amount of network traffic required for starting a Prefect flow run:

Our analysis shows that pulling packages accounts for 76% of container start time, but only 6.4% of that data is read.

Additionally, update docs for using the eStargz format with customer image builds.

Example Use

https://github.com/moby/buildkit?tab=readme-ov-file#imageregistry

oci-mediatypes=true: use OCI mediatypes in configuration JSON instead of Docker's
compression=<uncompressed|gzip|estargz|zstd>: choose compression type for layers newly created and cached, gzip is default value. estargz should be used with oci-mediatypes=true.
compression-level=<value>: compression level for gzip, estargz (0-9) and zstd (0-22)

Additional context

• Introducing Seekable OCI for lazy loading container images

• AWS Fargate enables faster container startup using Seekable OCI

• Slacker: Fast Distribution with Lazy Docker Containers

• Building containers without waiting for pull completion of base images on BuildKit

• https://github.com/google/crfs

• https://github.com/containerd/stargz-snapshotter