COPY . . command in Dockerfile for ASP.NET

Question

The Visual Studio tooling for Docker creates a Dockerfile for ASP.NET projects containing a COPY . . command as below:

WORKDIR /src
COPY *.sln ./
...
COPY . .

From what I've read, the <src> parameter is relative to the context, so isn't affected by the WORKDIR /src command. The <dest> however is relative to the WORKDIR so will be pointing at /src.

Is this command just bringing over the remaining files from the root for packaging (docker-compose.yml, .dockerignore, etc.)? If so, then why is this done ahead of the RUN dotnet build... command?

Full Dockerfile below:

FROM microsoft/aspnetcore:2.0 AS base
WORKDIR /app
EXPOSE 80

FROM microsoft/aspnetcore-build:2.0 AS build
WORKDIR /src
COPY *.sln ./
COPY MyProject/MyProject.csproj MyProject/
RUN dotnet restore
COPY . . # The line mentioned above
WORKDIR /src/MyProject
RUN dotnet build -c Release -o /app

FROM build AS publish
RUN dotnet publish -c Release -o /app

FROM base AS final
WORKDIR /app
COPY --from=publish /app .
ENTRYPOINT ["dotnet", "MyProject.dll"]

Answer 1

Some more pointers on the above from Scott Hanselman: https://www.hanselman.com/blog/OptimizingASPNETCoreDockerImageSizes.aspx

PRO TIP: Docker is smart about making intermediate images and doing the least work, but it's useful if we (the authors) do the right thing as well to help it out.

For example, see where we COPY the .csproj over and then do a "dotnet restore"? Often you'll see folks do a "COPY . ." and then do a restore. That doesn't allow Docker to detect what's changed and you'll end up paying for the restore on EVERY BUILD.

By making this two steps - copy the project, restore, copy the code, this means your "dotnet restore" intermediate step will be cached by Docker and things will be WAY faster.

Answer 2

The COPY . . copies the entire project, recursively into the container for the build.

The reason for the separation of the first 2 COPY commands with dotnet restore and then the complete COPY with dotnet build is a Docker caching trick to speed up builds. It is done this way so the project dependencies don't need to be installed every time a code change is made.

Docker images are built in layers. Docker compares the contents and instructions that would make up the each new layer to previous builds. If they match the SHA256 checksum for the existing layer, the build step for that layer can be skipped.

Code changes a lot more than dependencies, and dependencies are usually fetched from a slow(ish) network now. If you copy the code after the dependencies are completed then you don't bust the cached dependency layer for every other change.

The Node.js equivalent does the package.json before the app contents:

WORKDIR /app
COPY package.json /app/
RUN npm install
COPY . /app/
CMD ["node", "app/index.js"]