Docker max depth exceeded error occurs when we use “build” and “image” at the same time and with the Dockerfile.

As part of our Docker hosting support service, Bobcares responds to all inquiries, big or small.

Let’s take a closer look at how our Support team handled the error: Service ‘api’ failed to build: max depth exceeded

Docker max depth exceeded: How to solve?

To resolve the error,we just need to execute docker system prune -a to remove any stopped container. This command deleted all the local docker images related to the dockerfile.

To fix the error, simply run docker system prune -a to remove any stopped containers. This command removed all local Docker images associated with the Dockerfile. The local storage has reached a limit after so many builds, and thus the error max depth has been exceeded.

As the number of images increases, so does the local storage, resulting in the error: max depth exceeded. The maximum depth does not indicate a storage-capacity error. Instead, it indicates that the api image we were creating had too many layers.

One plausible explanation is that we have a recursion as a result of having this in the compose file:


image: mhart/alpine-node:12
build: ./client

as well as this in a Dockerfile
FROM mhart/alpine-node:12

Every time we run our build, it adds a few layers to the local mhart/alpine-node:12 image. Running docker history mhart/alpine-node:12 confirms this. If this is the case, we should rename the image in the compose file.

Image parameter

In a docker-compose.yml definition, the image parameter for a service has two meanings, depending on whether or not a build parameter is present.

• The image will be pulled and run if there is no build stanza.
• If a build stanza is present, image will be the name of the built image that is tagged as, and run.

By naming the built image microsoft/mssql-server-linux, which corresponds to the FROM image microsoft/mssql-server-linux. Every time, Docker stacked the build on top of itself.

The initial build used the “official” microsoft/mssql-server-linux image, but subsequent builds used the local microsoft/mssql-server-linux image that had been appended to, until we reached the storage driver’s maximum number of layers.

For all images we create, use our own namespace:
version: "3"
services:
mssql:
build: .
image: 'user3437721/mssql-server-linux'


How to fix?

We’ll get the error “max depth exceeded” if we use “build” and “image” in the Dockerfile at the same time. Make sure we’re using a different image name than the one our Dockerfile starts with to solve this problem. Otherwise, each build will add layers to our “final image,” bringing us closer to the maximum depth.

Another consequence of this issue is that docker builds will always start from the beginning as if we had requested that the cache be disabled. This is because we’re updating the initial image with each build. Simply put, it does not update the image from which our build is based. The image builds will resume using the cache if we fix the docker-compose.yml file, which will be much faster.

[Looking for a solution to another query? We are just a click away.]

Conclusion

To sum up, our Support team handled the error: Service ‘api’ failed to build: max depth exceeded