Unable to build our Docker image

Alexandre Fillatre

In our first pipeline step, we're calling `yarn install`, but this fails right away : 

 

<span>+ yarn install</span>

# node[39]: std::unique_ptr<long unsigned int> node::WorkerThreadsTaskRunner::DelayedTaskScheduler::Start() at ../src/node_platform.cc:68

# Assertion failed: (0) == (uv_thread_create(t.get(), start_thread, this))

----- Native stack trace -----

1: 0xca5580 node::Abort() [node]
2: 0xca5600 node::errors::SetPrepareStackTraceCallback(v8::FunctionCallbackInfo<v8::Value> const&) [node]
3: 0xd24949 node::WorkerThreadsTaskRunner::WorkerThreadsTaskRunner(int) [node]
4: 0xd24a6c node::NodePlatform::NodePlatform(int, v8::TracingController*, v8::PageAllocator*) [node]
5: 0xc5e5d3 [node]
6: 0xc5ef64 node::Start(int, char**) [node]
7: 0x7f23b8cd424a [/lib/x86_64-linux-gnu/libc.so.6]
8: 0x7f23b8cd4305 __libc_start_main [/lib/x86_64-linux-gnu/libc.so.6]
9: 0xbb8eee _start [node]

<span><br><br>Here's my bitbucket-pipeline :</span>

 

image: google/cloud-sdk:latest

definitions:
services:
docker:
# See https://community.atlassian.com/t5/Bitbucket-questions/Pipeline-broken-after-update-to-v25-0-2-multiarch-prod-stable/qaq-p/2597310#U2598724
      image: docker-public.packages.atlassian.com/sox/atlassian/bitbucket-pipelines-docker-daemon:v20.10.24-multiarch-runc-patch-prod-stable
steps:
    - step: &build-docker-staging
        runs-on:
          - self.hosted
          - gcp
image: node:20
size: 2x
caches:
          - node
name: Build and push docker image
script:
          - yarn install
services:
          - docker

pipelines:
branches:
master:
      - step: *build-docker-staging

 

<span>Using node:20.2 as base image fixes the issue, but we need a newer version.</span>

 

<span>Right now, I have no idea why the build doesn't work, and how to fix that.</span>

 

<span>Thanks</span>

Syahrul

G'day, @Alexandre Fillatre 

Could you raise a support ticket with us on our Suppor portal so we can investigate this issue further?

Also, before you raise a support ticket with us, please add --verbose options on your yarn install line in YAML, as they should give us more verbose output that will help the investigation.

Regards,
Syahrul