ERROR: error executing step command 'deploy --all': failed deploying service 'backend': archive/tar: write too long

[Hyunduck-Hwang]

Hello,
First of all, I want to thank about RAG deep dive series and this repo.

I have an issue. The error message is the same as title.
ERROR: error executing step command 'deploy --all': failed deploying service 'backend': archive/tar: write too long

I followed the instruction with doc in here.

azd auth login
azd env new

azd env set AZURE_AUTH_TENANT_ID {my-tenant-id}
azd env set AZURE_RESOURCE_GROUP skprac
...

After I run, azd up. All seems good untill deploying backend.

Creating Python virtual environment "app/backend/.venv"...
Installing dependencies from "requirements.txt" into virtual environment (in quiet mode)...
Updating redirect URIs for client app ID e4df05e1-ac0c-490f-bef5-3f45596d9b68...
Application update for client app id e4df05e1-ac0c-490f-bef5-3f45596d9b68 complete.
Creating Python virtual environment "app/backend/.venv"...
Installing dependencies from "requirements.txt" into virtual environment (in quiet mode)...
Running "prepdocs.py"
[06:41:27] INFO     Loading azd env from /workspaces/azure-search-openai-demo/.azure/fed/.env, which may override existing environment variables                                                                                                                               load_azd_env.py:28
           INFO     Connecting to Azure services using the azd credential for tenant                                                                                                                                                         prepdocs.py:321
           INFO     Using local files: ./data/*                                                                                                                                                                                                                                   prepdocs.py:104
           INFO     Setting up search index using integrated vectorization...                                                                                                                                                                                 integratedvectorizerstrategy.py:121
           INFO     Checking whether search index gptkbindex exists...                                                                                                                                                                                                        searchmanager.py:70
[06:41:29] INFO     Creating new search index gptkbindex                                                                                                                                                                                                                      searchmanager.py:75
           INFO     Including parent_id field in new index gptkbindex                                                                                                                                                                                                        searchmanager.py:142
           INFO     Including vectorizer for search index gptkbindex, using Azure OpenAI service cog-2paaqmzkbawl6                                                                                                                                                           searchmanager.py:162
[06:41:35] INFO     Skipping ./data/Shrub's Paradox - by Le Shrub.pdf, no changes detected.                                                                                                                                                                               listfilestrategy.py:104
           INFO     Skipping ./data/paradox.txt, no changes detected.                                                                                                                                                                                            listfilestrategy.py:104
[06:41:37] INFO     Successfully created index, indexer: gptkbindex-indexer, and skillset. Please navigate to search service in Azure Portal to view the status of the indexer.                                                                               integratedvectorizerstrategy.py:185

Deploying services (azd deploy)

  (x) Failed: Deploying service backend

ERROR: error executing step command 'deploy --all': failed deploying service 'backend': archive/tar: write too long

I am stuck with this error for 2 days. I think I cannot solve this issue with myself so here to ask some help!!
Thanks!

[pamelafox]

Can you provide more details on your environment: What OS? What's your Python version?
I assume this is deploying to Azure Container Apps?
It looks like this is an error coming from the azd CLI, as that error message is associated with the Go programming language.
We may want to report it here: https://github.com/Azure/azure-dev/issues

[vhvb1989]

most likely a very long (more than 100 char) path. Can you confirm the length of the path where you init the project in your system file?

[Hyunduck-Hwang]

@pamelafox I deleted the container and image, then reinstalled it. Now it works like a charm. Sorry for bothering. I was using WSL2(ubuntu) and Azure Container Apps. I don't know about python version. It maybe 3.11.11
I will let you know with extra info on my environments when the error occurs again !!
Should I close this issue?

@vhvb1989 I get the same cause when I asked this error to claude sonnet. But it seems not the main reason idk...Anyway thanks for helping!

vscode ➜ /workspaces/azure-search-openai-demo (main) $ pwd
/workspaces/azure-search-openai-demo

[eavestn]

@Hyunduck-Hwang what was the cause? Where did you delete the container and image? Docker? I am asking because I am having the same issue.

[vhvb1989]

@weikanglim , could this issue be related to: Azure/azure-dev#4773 ?

I wonder if that fix would side-fix and help here.

[Hyunduck-Hwang]

@eavestn Sorry to hear you having the same issue. I don't know what the cause was. I am using docker desktop since I am not ML ops/engineer(More of a ML researcher). You can find the bin/trash can icon at the right side of docker desktop when you click containers and images tab left side. However, It seems that is not enough. Also I deleted the azure-search-openai-demo container on VScode. You can find it at the left panel when you open VScode. There is a monitor icon called Remote Explorer. After you enter remote explorer, you need to click Dev containers. After I deleted the container and the image, I repeated installation all over again. And it works suddenly....GL

[pamelafox]

I've also gotten this when I accidentally had a .venv inside app/backend folder, that made the container too large. Make sure you dont have any additional folders auto-added that dont need to be in the Docker image.

Closing as I don't think there's anything to do on our side.

[leongj]

I just had the exact same issue -- because I had backend_env folder in /app/backend. Deleted it and it now deploys