adding pdfcpu as go dependency increases container size in 600mb

[glserranv]

Adding pdfcpu as a dependency increases 600MB to container size. Is this exepected/known?

[hhrutter]

Hello!

I don't have a good explanation for this but
if you add pdfcpu as a dependency in your Go app this may be the case.
This may be related to pdfcpu's dependency - see go.mod

[hhrutter]

Can you investigate further?
Also is your container size increasing by or to 600MB?

[J-Rocke]

I can build docker container with less than 70MB including pdfcpu as a go library and many more libs. I only include the binary and not the source code in the container which is a common best practice. I don't think this is a issue with pdfcpu but with the way the container is build.

[hhrutter]

👍 good news!

Is your docker file any different from the one in the repo, if not would you mind sharing yours?

# Dockerfile References: https://docs.docker.com/engine/reference/builder/

# Start from the latest golang base image
FROM golang:latest as builder

# install
RUN go install github.com/pdfcpu/pdfcpu/cmd/pdfcpu@latest

######## Start a new stage from scratch #######

FROM alpine:latest

RUN apk --no-cache add ca-certificates

WORKDIR /root/

# Copy the Pre-built binary file from the previous stage
COPY --from=builder /go/bin .

# Command to run the executable
CMD ["./pdfcpu"]

[J-Rocke]

@hhrutter I have a very similar Dockerfile but in our usecase pdfcpu is called directly from the go code. Since there is no evidence that pdfcpu greatly inflates binary size / container size this issue is missleading in my opinion and should maybe be closed. If you want to investigate more you can of course keep it open.

[hhrutter]

sounds good to me.