Stack: Docker build consistently runs out of memory

[dpyro]

General summary/comments (optional)

Steps to reproduce

1. Add to stack.yaml:

docker:
  enable: true

1. Run

stack build

Expected

Stack to build project using a docker instance.
Stack to build project in a reasonable timeframe.

Actual

The docker image runs out of memory while GHC is compiling dependencies. The project will eventually finish compiling if you run stack build over and over again--there will be less packages to complete for each invocation. There does not seem to be an obvious way to compile these packages beforehand or to use a binary version which makes the save-compile-test cycle take a very long time on even a modern machine.

$ stack build
[...]
tls-1.4.0: copy/register
vector-0.12.0.1: copy/register
Progress: 93/121
--  While building custom Setup.hs for package Cabal-2.0.1.1 using:
      /Users/dpyro/.stack/setup-exe-cache/x86_64-linux-dkda49f7ca9b244180d3cfb1987cbc9743/Cabal-simple_mPHDZzAJ_2.0.1.0_ghc-8.2.2 --builddir=.stack-work/dist/x86_64-linux-dkda49f7ca9b244180d3cfb1987cbc9743/Cabal-2.0.1.0 build --ghc-options " -ddump-hi -ddump-to-file -fdiagnostics-color=always"
    Process exited with code: ExitFailure (-9) (THIS MAY INDICATE OUT OF MEMORY)
    Logs have been written to: /Users/dpyro/Projects/dexer/.stack-work/logs/Cabal-2.0.1.1.log

    Configuring Cabal-2.0.1.1...
    Preprocessing library for Cabal-2.0.1.1..
    Building library for Cabal-2.0.1.1..
    [  1 of 168] Compiling Distribution.Compat.Binary ( Distribution/Compat/Binary.hs, .stack-work/dist/x86_64-linux-dkda49f7ca9b244180d3cfb1987cbc9743/Cabal-2.0.1.0/build/Distribution/Compat/Binary.o )
    [  2 of 168] Compiling Distribution.Compat.Exception ( Distribution/Compat/Exception.hs, .stack-work/dist/x86_64-linux-dkda49f7ca9b244180d3cfb1987cbc9743/Cabal-2.0.1.0/build/Distribution/Compat/Exception.o )
[...]

Stack version

$ stack --version
Version 1.6.5 x86_64 hpack-0.20.0

Method of installation

• brew install stack

[borsboom]

How is your Docker Engine configured? If it's running in a VM (e.g. when using Docker Machine or Docker for Mac), have you checked that it has enough memory allocated?

[pingiun]

I had this problem (with docker for mac), and setting the preferences so docker uses more memory fixed it for me.

[dbaynard]

Thanks, @pinguin. I'm going to close, then. @dpyro (or anybody else), do reopen if that doesn't solve your problem.

[benjamingoldstein]

To solve this, I was initially trying docker run --memory=4g ... but it didn't change anything. I had to go to the docker icon in the top bar (on Mac) and go to "Preferences" and then "Resources", and then slide the slider up to 4gb.