search

Azure-Samples / agent-openai-python-prompty-langchain-pinecone

Function calling for vector database lookup based on user question.
MIT License
22 stars 11 forks source link

azd up failed when using eastus as location/region #44

Closed puicchan closed 1 month ago

puicchan commented 1 month ago

Please provide us with the following information:

This issue is for a: (mark with an x)

- [ ] bug report -> please search issues before submitting
- [ ] feature request
- [x] documentation issue or request
- [ ] regression (a behavior that used to work and stopped in a new release)

Minimal steps to reproduce

  1. azd init -t agent-openai-python-prompty-langchain-pinecone
  2. azd auth login
  3. azd env set PINECONE_API_KEY ***
  4. azd up
  5. select eastus for location

I picked eastus because of this line in README "We recommend using swedencentralfor Azure OpenAI and eastus for the speech to text services". azd up is successful when I use eastus2. Should we consider recommending a better region?

Any log messages given by the failure

ERROR: error executing step command 'provision': deployment failed: error deploying infrastructure: deploying to subscription:

Deployment Error Details: InvalidTemplateDeployment: The template deployment 'cognitiveServices' is not valid according to the validation procedure. The tracking id is 'cf1383ad-6627-4692-bb08-8f8d3b7a6be3'. See inner errors for details. InsufficientQuota: This operation require 20 new capacity in quota Tokens Per Minute (thousands) - Text-Embedding-Ada-002, which is bigger than the current available capacity 15. The current quota usage is 465 and the quota limit is 480 for quota Tokens Per Minute (thousands) - Text-Embedding-Ada-002.

TraceID: 650ff4dc66db539ea7a093c76ad8d0f9

Expected/desired behavior

OS and Version?

Windows 10.

Versions

azd version 1.9.5

Mention any other details that might be useful

@kristenwomack - FYI

Thanks! We'll be in touch soon.

cassiebreviu commented 1 month ago

@kristenwomack This a quota issue, not an issue with the template itself.

kristenwomack commented 1 month ago

The template says suggests using eastus but isn't working in eastus because of quota (is that the region quota or the region + sub combo quota or just sub quota)? PC was able to deploy to eastus2 so should the template README suggest eastus2?

cassiebreviu commented 1 month ago

Quota is specific to each sub. Model availability is not. So the quota issues are based on the individual sub and you would need to go to the quota mgmt screen and find what sub has the model you need in the region you have with quota.