I'm trying to improve the efficiency of my Docker builds for a Go project, but I can't seem to get the caching to work properly. I'm using the following Dockerfile:
```
FROM golang:1.24 AS builder
WORKDIR /workspace
COPY go.mod go.mod
COPY go.sum go.sum
RUN go mod download
# Copy the Go source (relies on .dockerignore to filter)
COPY . .
ENV GOCACHE=/root/.cache/go-build
RUN --mount=type=cache,target=/root/.cache/go-build
CGO_ENABLED=0 go build -a -o manager cmd/main.go
FROM gcr.io/distroless/static:nonroot
WORKDIR /
COPY --from=builder /workspace/manager .
ENTRYPOINT ["/manager"]
```
When I run `docker build .`, I notice that the 'RUN go mod download' layer is cached, which is great. However, the 'RUN --mount=type=cache,target=/root/.cache/go-build CGO_ENABLED=0 go build -a -o manager cmd/main.go' always takes over 2 minutes to execute, even though the code hasn't changed. I thought using the cache mount would help speed up the build process. What am I doing wrong, and how can I make sure these Go builds are properly cached?
1 Answer
You might want to consider using multistage builds more effectively by caching layers instead of relying solely on the mount. If the Go source code hasn’t changed much, you can cache your dependencies more efficiently by structuring your Dockerfile to keep the go.mod and go.sum part separate from the rest of your builds.
Can you elaborate on how to implement that? I see you're already caching the go mod download, which helps a lot. However, I’m concerned about the recompilation step, since most of my Go code stays consistent. I feel like there should be a way to cache those builds as well.