Unable to upload RAG context document using Search service through devops

Question

Unable to upload RAG context document using Search service through devops

Romel Costelo 0

Its constantly stuck at uploading state.

Aditya N 2,550 Reputation points Microsoft External Staff Moderator

2026-04-08T11:25:44.1633333+00:00
Hi @Romel Costelo

Please could you provide few more details.

What method are you suing for upload?

What is the size of document?

Any warnings or timeouts messages you encountered?

Did you see any error messages?
Romel Costelo 0 Reputation points

2026-04-09T06:08:16.61+00:00
Hi Aditya,

I am using Azure devops AI Work Item Assistant extension on Azure devops

Document is less than 500KB

No timeouts, just stuck in uploading state

No error message
Golla Venkata Pavani 3,925 Reputation points Microsoft External Staff Moderator

2026-04-18T00:19:51.89+00:00

Hi @Romel Costelo

Following up - Please could you let us know whether you had a chance to review our suggestion.
Ghulam Muhayyu Din 0 Reputation points

2026-04-19T17:29:04.6+00:00

Hello Romel,

When integrating Retrieval-Augmented Generation (RAG) systems into deployment pipelines, a document stuck in the "Uploading" state within the Azure DevOps AI Work Item Assistant usually indicates a silent failure in the Search service connection.

The UI unfortunately does not surface Azure AI Search indexer errors directly. This typically happens if an admin API key is missing or invalid, or if the target index is misconfigured. Validate your Search service endpoint and ensure you are passing a valid admin API key. If the configuration is correct, delete and recreate the target index. Recreating it will clear the stalled indexing pipeline and usually resolves the hang before you retry the upload.
Romel Costelo 0 Reputation points

2026-04-20T13:57:08.8066667+00:00

Thank you for your comments. We are now able to use the upload feature.

I have a question—does this process store an actual file in the storage account? We can see that the indexing is being completed successfully, but we’re unable to locate where the search service saves the data.
Golla Venkata Pavani 3,925 Reputation points Microsoft External Staff Moderator

2026-04-20T18:59:59.4533333+00:00
Hi @Romel Costelo

Thank you for the update, and I’m glad to hear the upload feature is now working.

Azure AI Search does not store the original file (for example, PDF or DOCX) in a storage account by default.

The original documents remain in the source system (such as Azure Blob Storage, if used as a data source), and the search service only reads from that source during indexing.

During ingestion, Azure AI Search extracts text and metadata from the document and stores it in a search index, which is optimized for querying.

The search index is stored within the Azure AI Search service itself, and once populated, it exists independently of the original source data

Reference:
https://learn.microsoft.com/en-us/azure/search/search-indexer-overview
https://learn.microsoft.com/en-us/azure/search/search-how-to-index-azure-blob-storage
https://learn.microsoft.com/en-us/azure/search/search-blob-storage-integration

Kindly let us know if the above helps or you need further assistance on this issue.

Please "upvote" if the information helped you. This will help us and others in the community as well.
Golla Venkata Pavani 3,925 Reputation points Microsoft External Staff Moderator

2026-04-21T19:40:42.89+00:00

Hi @Romel Costelo

I'm just reaching out to see if your issue has been resolved or if you've had a chance to review my previous comment?

1 answer

Your answer

Aditya N 2,550 Reputation points Microsoft External Staff Moderator

2026-04-08T11:25:44.1633333+00:00

Hi @Romel Costelo

Please could you provide few more details.

What method are you suing for upload?

What is the size of document?

Any warnings or timeouts messages you encountered?

Did you see any error messages?
Romel Costelo 0 Reputation points

2026-04-09T06:08:16.61+00:00

Hi Aditya,

I am using Azure devops AI Work Item Assistant extension on Azure devops

Document is less than 500KB

No timeouts, just stuck in uploading state

No error message
Golla Venkata Pavani 3,925 Reputation points Microsoft External Staff Moderator

2026-04-18T00:19:51.89+00:00

Hi @Romel Costelo

Following up - Please could you let us know whether you had a chance to review our suggestion.
Ghulam Muhayyu Din 0 Reputation points

2026-04-19T17:29:04.6+00:00

Hello Romel,

When integrating Retrieval-Augmented Generation (RAG) systems into deployment pipelines, a document stuck in the "Uploading" state within the Azure DevOps AI Work Item Assistant usually indicates a silent failure in the Search service connection.

The UI unfortunately does not surface Azure AI Search indexer errors directly. This typically happens if an admin API key is missing or invalid, or if the target index is misconfigured. Validate your Search service endpoint and ensure you are passing a valid admin API key. If the configuration is correct, delete and recreate the target index. Recreating it will clear the stalled indexing pipeline and usually resolves the hang before you retry the upload.
Romel Costelo 0 Reputation points

2026-04-20T13:57:08.8066667+00:00

Thank you for your comments. We are now able to use the upload feature.

I have a question—does this process store an actual file in the storage account? We can see that the indexing is being completed successfully, but we’re unable to locate where the search service saves the data.
Golla Venkata Pavani 3,925 Reputation points Microsoft External Staff Moderator

2026-04-20T18:59:59.4533333+00:00

Hi @Romel Costelo

Thank you for the update, and I’m glad to hear the upload feature is now working.

Azure AI Search does not store the original file (for example, PDF or DOCX) in a storage account by default.

The original documents remain in the source system (such as Azure Blob Storage, if used as a data source), and the search service only reads from that source during indexing.

During ingestion, Azure AI Search extracts text and metadata from the document and stores it in a search index, which is optimized for querying.

The search index is stored within the Azure AI Search service itself, and once populated, it exists independently of the original source data

Reference:
https://learn.microsoft.com/en-us/azure/search/search-indexer-overview
https://learn.microsoft.com/en-us/azure/search/search-how-to-index-azure-blob-storage
https://learn.microsoft.com/en-us/azure/search/search-blob-storage-integration

Kindly let us know if the above helps or you need further assistance on this issue.

Please "upvote" if the information helped you. This will help us and others in the community as well.
Golla Venkata Pavani 3,925 Reputation points Microsoft External Staff Moderator

2026-04-21T19:40:42.89+00:00

Hi @Romel Costelo

I'm just reaching out to see if your issue has been resolved or if you've had a chance to review my previous comment?

Answer 1

Hi @Romel Costelo

Thank you for providing answers.

When uploading a RAG context document using Azure AI Search through the Azure DevOps AI Work Item Assistant, the upload can remain stuck in an “Uploading” state with no error message.

Why this happens: Because the AI Work Item Assistant does not surface Azure AI Search indexer status or errors in the UI, the upload can appear to remain indefinitely in “Uploading” if:

An admin API key is not used

The index is misconfigured or unavailable

Resolution Steps:

Validate Azure AI Search configuration

Confirm the Search service endpoint and admin API key are correct. Verify the Azure AI Search service is running and reachable.

Recreate the Azure AI Search index

Delete and recreate the index configured for RAG context in the AI Work Item Assistant settings. Recreating the index resets stalled or invalid indexing pipelines.

Retry document upload

After confirming Search service configuration and index health, retry the upload.

Review Azure AI Search diagnostics

Indexing errors and failures are observable only through Azure AI Search diagnostics and logs, not through the Azure DevOps extension UI.

Reference: https://learn.microsoft.com/azure/search/search-security-api-keys https://learn.microsoft.com/azure/search/search-indexer-overview https://learn.microsoft.com/azure/search/search-indexer-overview#how-indexers-work

If the answer is helpful, Please do click "Accept the answer” and Yes, this can be beneficial to other community members.

If you have any other questions, let me know in the "comments" and I would be happy to help you

Share via

Unable to upload RAG context document using Search service through devops

1 answer

Your answer