This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(300.8, 41%, 39%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EMS%3C/text%3E%3C/svg%3E)
Hi
I have gone through the lab exercise mslearn-knowledge-mining as prescribed here: https://microsoftlearning.github.io/mslearn-knowledge-mining/Instructions/Exercises/01-azure-search.html
When i try to index the documents, i follow all the steps. Everytime it fails creating margies-indexer and leaves me with:
Document key:
localId=https%3a%2f%2fmystorage.blob.core.windows.net%2fmargies%2fcollateral%2fDubai%2520Brochure.pdf&documentKey=https%3a%2f%2fmystorage.blob.core.windows.net%2fmargies%2fcollateral%2fDubai%2520Brochure.pdf
Operation:
Enrichment.OcrSkill.#5
Message:
Could not execute skill.
Details:
An unexpected error occurred.
Does anybody know how to solve this?
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(224.00000000000003, 86%, 31%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EA%3C/text%3E%3C/svg%3E)
Welcome to the Microsoft Q&A forum.
Thank you for bringing this issue to our attention. We will update you as soon as possible.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(12.8, 70%, 11%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ES%3C/text%3E%3C/svg%3E)
Thank you for reaching out to us on Microsoft Q&A forum.
The error you are encountering with the OCR skill indicates that there might be an issue with the OCR processing during the indexing.
Here are a few steps you can take to troubleshoot and resolve this issue:
1.Make sure the documents in your Blob Storage are in a format supported by the OCR skill, such as PDF, TIFF, or JPEG. Check that the documents are not corrupted and can be opened manually.
2.Verify that your Azure AI Services resource is properly configured with enough quota for OCR operations. Ensure the pricing tier and region support OCR.
3.Log into the Azure Portal, navigate to the Cognitive Services resource, and check for any service-related issues or quota limitations.
4.Verify the connection string and access permissions for Azure Blob Storage are correctly set and ensure the Azure AI Search resource has access to the Azure AI Services resource.
5.Ensure the skillset configuration, especially for OCR, is correct and references the fields and parameters accurately. Check the Azure AI Search resource logs and indexer execution details for detailed error messages.
6.Try using a different document to check if the problem continues. Use a simple text file to rule out specific content issues. Also, update your configurations, save, and try indexing again to resolve intermittent problems.
If the issue persists, please follow the link and try the steps mentioned there,
Error: Could not execute skill->https://learn.microsoft.com/en-us/azure/search/cognitive-search-common-errors-warnings#error-could-not-execute-skill
Please attempt the exercise in the updated document: https://microsoftlearning.github.io/AI-102-AIEngineer/Instructions/22-azure-search.html#index-the-documents
Additional link: https://learn.microsoft.com/en-us/azure/search/search-create-service-portal
Please don't hesitate to reach out to us if you have any other queries.
If you found the information helpful, we would greatly appreciate it if you could acknowledge it by selecting the Accept Answer & Upvote options.
Hope you are doing well.
I wanted to check if you had the opportunity to review the information which was provided in in my previous posted answer.
Please feel free to contact us if you have any more quires.
If you feel that your quires have been resolved, please accept the answer by clicking the "Upvote" and "Accept Answer" on the post.
Thank you.
Hi Siddesh
So I've tried your suggestions, but the issue seems to persist. I did try to skip the OCR skillset, and in that case the same issue arrised with for example Enrichment.LanguageDetectionSkill.#3
I've also tried the same instructions as in the lab, but with plain text files. Here I got the same issue.
I have tested a bunch of different combinations of settings, and it seems to be my AI service that gives the problem.
Using the free ai service undet "Attach AI service" i am able to make an indexer.
When creating the AI services instance, i set location to West US 2, like my other resources, and pricing to Standard S0.
What can the issue with the AI service be?
Hi
The issue with your AI service instance might stem from a few potential factors
1.When setting up your AI service instance in West US 2 with Standard S0 pricing, ensure you review service quotas and configuration details. The Standard S0 tier imposes resource limitations, such as maximum transactions per second and concurrent request capacities. Exceeding these limits can lead to service errors, particularly for intensive tasks like OCR/language detection. Checking and adjusting your usage against these limits can help maintain optimal service performance.
2.Sometimes, even when using services in the same region, network latency or regional-specific issues can impact the performance of AI services, especially for complex tasks. Additionally, Azure regions might occasionally experience temporary outages, affecting service availability and performance. Checking the Azure Service Health Dashboard can provide updates on any ongoing issues affecting your region.
3.Make sure all settings for your AI service instance are correct, including endpoints, authentication keys, and resource connections. Misconfigurations can cause connectivity problems / service errors.
4.Make sure your Azure subscription is active and configured correctly for the chosen pricing tier. Billing problems / subscription restrictions can affect service availability.
Given that switching to the free AI service instance resolved your indexing problem, it suggests that the issue may indeed lie with the specific configuration or availability of the paid AI service instance. Continuing to monitor service health updates and reviewing configuration settings are essential steps to resolving such issues.
If you feel that your quires have been resolved, please accept the answer by clicking the "Upvote" and "Accept Answer" on the post.
Thank you.
Do you have a sample of how the configuration should be? I am following the instructions in https://microsoftlearning.github.io/AI-102-AIEngineer/Instructions/22-azure-search.html#index-the-documents. I can look for an error in configuration all i want, but if i dont know which config is required, i can really spot the mistake. I have checked the limitations and it should be fine.
Hi
To configure Azure AI Search for indexing documents, especially with skills like OCR and language detection, you typically need to define an indexer that specifies how your data will be pulled in and processed.
Here's a basic example of how an indexer configuration might look:
{
"dataSourceName": "your-data-source-name",
"targetIndexName": "your-target-index-name",
"skillsetName": "your-skillset-name",
"parameters": {
"configuration": {
"dataToExtract": "contentAndMetadata"
}
},
"outputFieldMappings": [
{
"sourceFieldName": "/document/content",
"targetFieldName": "content"
}
]
}
Sure, here's a simplified breakdown:
1.The dataSourceName refers to your source of data in Azure AI Search, such as a Blob Storage container.
2.targetIndexName is the name of the index in Azure AI Search where the processed data will be stored.
3.skillsetName is the name of the skillset configured in Azure AI Search, which determines the cognitive tasks like OCR or language detection applied during indexing.
4.parameters include configuration details for the indexer.
For instance, dataToExtract specifies what content (such as content and metadata) should be extracted from the documents.
5.outputFieldMappings specify how the processed data should be mapped to fields in your target index.
For example, mapping the extracted content to a specific field like content.
Make sure your skillset (your-skillset-name) is properly configured with the required skills (OCR, LanguageDetection, etc.) and that your data source and index are correctly defined and accessible.
[Reference,
FYI-> Create an indexer in Azure AI Search:https://learn.microsoft.com/en-us/azure/search/search-howto-create-indexers?tabs=portal
https://learn.microsoft.com/en-us/azure/search/search-how-to-load-search-index?tabs=sdk-dotnet ]
Thank you.