improve error messaging for "Your socket connection to the server... will be closed"

[Alambre020]

How did you install the Amplify CLI?

npm

If applicable, what version of Node.js are you using?

v19.0.1

Amplify CLI Version

10.6.2

What operating system are you using?

Ubuntu

Did you make any manual changes to the cloud resources managed by Amplify? Please describe the changes made.

There is an existing override.ts but is the original from functional deployments.

Describe the bug

Initializing an existing project under my own account.
Removed the amplify team-provider-info.json file.
Getting error on amplify push:

**
:~/Documents/DeckAnalyzer$ amplify push
✖ There was an error pulling the backend environment devtwo.
🛑 Packaging overrides failed.
Command failed with exit code 1: yarn install

Resolution: There may be errors in your overrides file. If so, fix the errors and try again.
Learn more at: https://docs.amplify.aws/cli/project/troubleshooting/

Session Identifier: 23f96950-8041-44f9-8a30-2a76743c0494

**

Expected behavior

Expected project to initialize and deploy.

Reproduction steps

amplify init
delete team-provider-info.json
amplify push

Project Identifier

Project Identifier: caa1b95f02a190bcb0b4ee7bcb9e7b6f

Log output

# Put your logs below this line

Additional information

No response

Before submitting, please confirm:

• I have done my best to include a minimal, self-contained set of instructions for consistently reproducing the issue.
• I have removed any sensitive information from my code snippets and submission.

[josefaidt]

Hey @Alambre020 👋 thanks for raising this! The reproduction steps do not seem to follow the typical paths for pushing Amplify projects. What is the use case for deleting the team-provider-info.json file?

[svidgen]

Hey @josefaidt, I was actually working with @Alambre020 to try to get him spun up on a personal project, and this is where we got stuck. (We hit a few other CLI snags as well, but I'm chalking that up to forgetting what order to run initial CLI setup commands in for now.)

Removing the team-provider-info.json is recommended in our docs when trying to share projects with completely separate backend deployments (different accounts): https://docs.amplify.aws/cli/teams/shared/#sharing-projects-within-the-team

If you want to share a project publicly and open source your serverless infrastructure, you should remove or put the amplify/team-provider-info.json file in the .gitignore file.

Feel free to pull me to help repro this one, as I own the original app + source.

[josefaidt]

Hey @svidgen and @Alambre020 perhaps we can update the docs to suggest only adding it to gitignore rather than removing it entirely, or clarify why it would be suggested to remove it. In short, you will still need this file to work within the project, including push attempts.

When sharing projects, the team-provider-info.json can be safely removed/ignored before the person you are sharing it with runs amplify init to create the project in their account.

If you do not remove the team-provider-info.json the person you are sharing the project with cannot initialize the project in their account using the same environment name as the existing envs in this file. For example, if I created the dev environment in my AWS account and shared the project with you, you would not also be able to create a dev environment, and you would instead be presented with an "AccessDenied" error since the CLI will recognize the environment name and attempt to reference resources in my account.

Additionally, if you do not remove this file I would be able to safely initialize the project using a different environment name such as devtwo or myowndev

[svidgen]

When sharing projects, the team-provider-info.json can be safely removed/ignored before the person you are sharing it with runs amplify init to create the project in their account.

This was the intended use-case. Initially we tried to use dev as well — I don't actually recall what error we saw there. After changing it to devtwo, it appeared to start successfully deploying until it hit the overrides error.

[josefaidt]

Thanks for clarifying @svidgen ! What sort of overrides issues were you running into?

[svidgen]

@Alambre020 I chatted with @josefaidt and he was able to pull the repo and push it without issue. Hopefully we just mindlessly flipped some commands. Let's try this fresh and make sure to run things strictly in this order:

1. git clone
2. remove team-provider-info.json
3. amplify init
4. amplify push

@josefaidt Thanks for your time. @Alambre020 or I will post back when we've had a chance to retry this.

[svidgen]

@Alambre020 Do you have a non-Ubuntu machine you can try this on?

[Alambre020]

@svidgen, it hasn't been turned on in long time but i can try tomorrow

[Alambre020]

@svidgen @josefaidt , I was able to test on a windows machine. I got through the init successfully. The push failed.

" There was an error pulling the backend environment dev. "

It cannot find .aws\conifg

[josefaidt]

Hey @Alambre020 does the ~/.aws/config file exist? If not, can you create one with a profile reference and region? For example:

# ~/.aws/config
[default]
region=us-east-1

[Alambre020]

It seems like it's looking in the wrong folder??

[josefaidt]

Hey @Alambre020 I do not recognize that directory name, but after a bit of googling it seems like this is related to Cygwin? Are you using Cygwin by chance? Can you check the value of your HOME env var?

[josefaidt]

Hey @Alambre020 are you still experiencing this issue?

[Alambre020]

@josefaidt , yes sorry... Been away from this in a while and will have to check this week. Last i remember i tried on an old windows machine and still had issues.

[Alambre020]

@josefaidt, my home variable is C:\Users\Nick\AppData\Roaming\SPB_Data which does seem to have the file that the error says is missing:

I do not see any cygwin installed.

[josefaidt]

Hey @Alambre020 no worries about cygwin, I was doing some digging on this path to understand what app SPB_Data relates to, and it appeared related to Cygwin.

Additionally, there is a typo in that path (.asw -> .aws)

[josefaidt]

Hey @Alambre020 are you still experiencing this?

[Alambre020]

@josefaidt , yes. Sorry for the delay and thank you for continuing to check in. I will have time to look at this again this weekend. I will have to double check the patching... If it is the cgwin... What should I try first? I'm not really familiar with it or not sure what i would've installed it for.

[Alambre020]

@josefaidt @svidgen , after renaming the folder I did get farther in the process. It's possible I named it incorrectly but should this be a folder that the amplify process should create?

Now the process timed out:

[Alambre020]

@josefaidt when i run the push again it says there are no changes? so maybe the first pushed worked after renaming the .aws folder?

[josefaidt]

Hey @Alambre020 thanks for checking on that! That error seems to be coming in after the CloudFormation push was successful, where there is some project state and files that are being uploaded at the end. This error and subsequent push attempt leads me to believe the operation was likely successful and the connection was not closed properly.

I will mark this as a bug to improve the error handling and messaging here

https://github.com/aws-amplify/amplify-cli/blob/dev/packages/amplify-provider-awscloudformation/src/aws-utils/aws-s3.ts#L134

Despite the oddity it does appear to be successful.