Why getting 500 Error from endpoint computervision/retrieval:vectorizeImage in West Europe?

Question

We are using Computer Vision API from Azure AI Services

computervision/retrieval:vectorizeImage

"api-version": "2024-02-01"
"model-version": "2023-04-15"

with Free tier deployed in West Europe.

Today at around 11.00 CET it stopped working and respond with:

{"error":{"code":"InternalServerError","message":"Internal server error"}}

Calling a new deployment in a different location works.

Status of Azure services In West Europe suggest this service is up and running, but we are getting the same error also in different deployments (free tier) in West Europe.

Is this related to the Free tier perhaps?

Answer 1

The 500 Internal Server Error you are encountering with the Computer Vision API in the West Europe region could be due to a few reasons. While the Azure service status indicates that the service is operational, there can still be transient issues that affect specific deployments, especially in the Free tier. Here are some potential causes and considerations:

1. Service Limitations: The Free tier may have certain limitations or quotas that, when exceeded, could lead to errors. If your usage patterns have changed or if there are more requests than the Free tier can handle, this could trigger a 500 error.
2. Transient Issues: Sometimes, internal server errors can occur due to temporary issues within the Azure infrastructure. These can resolve themselves after a short period, so retrying the request may yield a different result.
3. Deployment Configuration: Ensure that your deployment configuration is correct and that there are no issues with the model version or API version you are using. If the model version is not compatible with the API version, it could lead to errors.
4. Regional Issues: Since you mentioned that a deployment in a different location works, it could indicate a regional issue specific to West Europe. Azure services can sometimes experience localized outages or performance issues.

If the problem persists, consider reaching out to Azure support for further assistance, as they can provide more detailed insights into the specific error and any underlying issues affecting your deployment.

---

References:

• Troubleshoot online endpoint deployment and scoring
• Common deployment errors