Error in 0.0.4 when running publish on Node alpine image

[Hilaryous]

Description

When running rover subgraph publish --schema=schema.graphql --name=[service name] --routing-url=[service url] [graph name]@dev

I see this error:

Error: spawnSync /usr/local/share/.config/yarn/global/node_modules/binary-install/bin/rover ENOENT
    at Object.spawnSync (internal/child_process.js:1067:20)
    at spawnSync (child_process.js:602:24)
    at Binary.run (/usr/local/share/.config/yarn/global/node_modules/binary-install/index.js:84:20)
    at run (/usr/local/share/.config/yarn/global/node_modules/@apollo/rover/binary.js:71:10)
    at Object.<anonymous> (/usr/local/share/.config/yarn/global/node_modules/@apollo/rover/run.js:4:1)
    at Module._compile (internal/modules/cjs/loader.js:1063:30)
    at Object.Module._extensions..js (internal/modules/cjs/loader.js:1092:10)
    at Module.load (internal/modules/cjs/loader.js:928:32)
    at Function.Module._load (internal/modules/cjs/loader.js:769:14)
    at Function.executeUserEntryPoint [as runMain] (internal/modules/run_main.js:72:12) {
  errno: -2,
  code: 'ENOENT',
  syscall: 'spawnSync /usr/local/share/.config/yarn/global/node_modules/binary-install/bin/rover',
  path: '/usr/local/share/.config/yarn/global/node_modules/binary-install/bin/rover',
  spawnargs: [
    'subgraph',
    'publish',
    '--schema=schema.graphql',
    '--name=[service name]',
    '--routing-url=[service url]',
    '[graph name]@dev'
  ]
}

Steps to reproduce

Run yarn global add @apollo/rover@0.0.4, then
rover subgraph publish --schema=schema.graphql --name=[service name] --routing-url=[service url] [graph name]@dev
using a node:lts-alpine image.

Expected result

I should be able publish the schema

Actual result

I cannot publish the schema

Environment

• Operating system and version: node:lts-alpine (node version 14.16.0, alpine version 3.11) Link to Dockerfile
• Shell (bash/zsh/powershell): bash
• Version of Rover (rover --help): 0.0.4
• Apollo Studio Plan: enterprise
• Installation method: yarn

Note

I see #375 about 0.0.4 and alpine, but this seemed to be a different issue.

[abernix]

I suspect that #375 is still relevant as ENOENT is also saying "No such file or directory". For example, this happens even just running rover --version after installing the @apollo/rover package.

As a suggestion you may find useful, you could leverage Docker multi-stage builds to run Rover using a different image while still getting your app into an Alpine Node image. Depending on where in your build you need the rover subgraph publish command to run, the actual form the Dockerfile would take on could vary.

(I realize this isn't the real solution for having this command just work in Alpine, but I'm willing to try and help craft a workaround if there's an existing Dockerfile you can share a snippet of.)

[EverlastingBugstopper]

Closing this in favor of #537