lockfile | Handle locking via pid files | File Utils library

As has been commented in the comments section, the issue has been solved by moving the project folder outside the OneDrive directory.

Because it's a synchronized cloud folder, as soon as you install/add/delete/update anything, it's being updated in the OneDrive cloud so the file/folder it's being used in the background and potentially unreachable. If at this time you try to develop the project (gatsby develop or gatsby build) and the file is being used, you won't be able to run it.

I don't think it's a good practice to use a cloud folder because the amount of data synchronized (mainly because of the node_modules) it's something to care about (it's also ignored in the .gitignore for a reason) so moving it to any other folder outside the OneDrive directory should be enough to run your project because the rest of global dependencies, according to your logs, were successfully installed.

COPY normally copies things into the image owned by root, and it will create directories inside the image if they don't exist. In particular, when you COPY ./api/package.json ./api/, it creates the api subdirectory owned by root, and when you later try to run yarn install, it can't create the node_modules subdirectory because you've switched users.

I'd recommend copying files into the container and running the build process as root. Don't chown anything; leave all of these files owned by root. Switch to an alternate USER only at the very end of the Dockerfile, where you declare the CMD. This means that the non-root user running the container won't be able to modify the code or libraries in the container, intentionally or otherwise, which is a generally good security practice.

This seems like everything is working correctly rails just likes to output what its doing to the console but it should have added those to your file.

Microsoft updated their VMs around the 3rd December.
In this change the default node version has changed from 14.x to 16.x.

We found this exact issue when running on node 16.

This won't be a Long term fix - As you'll need to work out why your build doesn't work on Node 16, but in the short term you can add a command which tells the build agent to use version 14 of node.

Presuming you are using the modern YAML build pipelines, try adding the following as the first step in your YAML pipeline:

Adding the suitable env vars solved my problem.

I get values and files only by get Artifacts via GitLab CI API

previous - create token for your User Gitlab and create var GITLAB_USER_TOKEN

Your rootfs for test-image is larger than 4GB, which is the maximum filesystem size for FAT partitions (the default used by the .hddimg image type).

The error already contains two solutions:

1. Reduce the image size, by removing packages
2. Use the wic format (vmdk and vdi are for virtual machine engines)

To use the wic format, simply add wic to IMAGE_FSTYPES and Yocto will build a .wic image and place it in the deploy/images directory. You can flash this to your disk using dd or bmaptool, and boot it.

These images are built from a template specified by WKS_FILE, which defaults to genericx86.wks.in for the the genericx86-64 machine. This default template creates a partition table and installs an EFI bootloader (Grub).

If you'd like to manually generate the wic images or modify the templates, check out the documentation for Creating Partitioned Images Using Wic.

It looks like your downstream pipeline is defining a workflow with 2 independent rules: if and changes. This may cause the jobs to be included if the first condition in the if is met, i.e. if it is a MR pipeline. Try removing the dash in front of changes, as in the example here, to treat this as a single rule:

That means that the package-lock.json file was created while performing an npm install with an npm version less than 7.

npm 7 has changed not only the format of package-lock.json, but also how it handles peer dependencies. When you upgrade to a newer npm version ( >=7 ) make sure to test everything properly with a freshly checked out repo that still has the old file version.

You can resolve the issue (when all testing went well) by committing and pushing the package-lock.json in the new format.