Image build fails if it takes longer than 20 minutes

[voroninman]

Community Note

• Please vote on this issue by adding a 👍 reaction to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment

Terraform (and docker Provider) Version

Terraform v1.0.3 on darwin_amd64

• provider registry.terraform.io/kreuzwerker/docker v2.14.0

Affected Resource(s)

• docker_image
• docker_registry_image

Expected Behaviour

The image is built by Terraform.

Actual Behaviour

...
module.foo.docker_registry_image.this["pytorch-gpu"]: Still creating... [19m50s elapsed]
╷
│ Error: Error building docker image: context deadline exceeded
│
│   with module.foo.docker_registry_image.this["pytorch-gpu"],
│   on ../main.tf line 153, in resource "docker_registry_image" "this":
│  153: resource "docker_registry_image" "this" {
│
╵

Steps to Reproduce

Try to build a Docker image in Terraform for the following Dockerfile:

FROM busybox:latest
RUN sleep 1201

Important Factoids

The timeout includes the uploading of the image.

Building a Docker image for a data science environment usually means downloading or compiling big software packages (Pytorch+CUDA in my case) that result in 2-6 Gb images. The time of building and uploading such a image sometimes exceeds 20 minutes.