azd cli time out when indexing many pdfs

[anarchitect]

Please provide us with the following information:

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

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

Minimal steps to reproduce

added my own PDFs (40+ files) and the indexing will take a while, and then run azd up, which then received this error on the terminal
AzureDeveloperCliCredential.get_token failed: {"type":"consoleMessage","timestamp":"2023-09-10T10:52:31.602493+10:00","data":{"message":"fetching token: failed to authenticate: server response error:\n Get "https://login.microsoftonline.com/common/discovery/instance?api-version=1.1\u0026authorization_endpoint=https%3A%2F%2Flogin.microsoftonline.com%xxxxxxxxxFauthorize\": net/http: TLS handshake timeout\n"}}

Any log messages given by the failure

please see error on the terminal above

Expected/desired behavior

it should index all documents without time out. I could not find a way to extend AZD credential time out

OS and Version?

macOS (13.5.1 (22G90))

azd version?

run azd version and copy paste here. 1.3.0

Versions

1.3.0

Mention any other details that might be useful

---

Thanks! We'll be in touch soon.

[pamelafox]

Interesting, I don't think I've run into azd credential timeouts before. @vhvb1989 Do we also need to refresh the credential the way we refresh openai credentials?

@anarchitect For now, could you remove prepdocs as a postprovision hook from azure.yaml, and then call ./scripts/prepdocs.sh after azd up completes? I think that might help with the timeout.

[github-actions]

This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this issue will be closed.

[pamelafox]

Thanks for reporting this and for the details. The timeout you hit appears to be a TLS handshake timeout to Azure AD during the azd postprovision hook, which is often environment/network related rather than specific to this sample.

Since this was opened in 2023 and there hasn’t been any activity since, we’re going to close it out. In the meantime, we’ve documented running data ingestion outside of azd up to avoid long-running postprovision steps:

• docs/azd.md explains the postprovision hook
• docs/data_ingestion.md shows how to run scripts/prepdocs.sh (or .ps1) manually after azd up

If you still encounter this with the latest azd and the current version of this sample, please try:

• Running prepdocs after azd up (remove the postprovision hook and call scripts/prepdocs.sh manually)
• Ensuring you’re logged in with azd auth login and have a stable network connection
• Providing your azd version and any logs/output

Feel free to open a new issue with that information and we’ll be happy to investigate further.