Azure AI Search Indexer Timeout Issue with PDF Document

Hi @Su Myat Hlaing ,

It seems like you're encountering a timeout issue with your Azure AI Search indexer when processing a PDF document. Here are some troubleshooting suggestions based on common issues related to indexer timeouts:

Ensure that your PDF document does not exceed the maximum size limit for document extraction, which is 134,217,728 bytes for your current service tier. If it does, the indexer will not be able to process it.

Verify that the content type of the PDF is supported. Unsupported content types will cause the indexer to skip the document.

Since you mentioned that this setup was working previously, it could be a transient issue. Consider waiting and trying to run the indexer again later. Setting your indexer to run on a schedule may help mitigate these transient errors.

If you're using a custom skill for processing the document, check if it is returning results consistently and ensure that it is not stuck in an infinite loop. The default timeout for a custom skill is 30 seconds, but you can increase this up to a maximum of 230 seconds if necessary.

If your custom skill processes multiple documents at once, consider reducing the batch size to ensure that it can execute within the timeout limits. Use the Azure portal or the REST API to check the indexer's status and see if there are any specific errors being reported.

Ensure that field mappings or change tracking values are not causing the document to be skipped. If the document was updated after the indexer ran, it may not appear in the search index until the next scheduled run.

Implementing these suggestions should help you identify and resolve the timeout issue with your indexer.

Error: Skill did not execute within the time limit
Missing documents
Transient errors

Hope this information is helpful, let me know if you have any further queries.

Hi @Bhargavi Naragani, Thank you for your detailed response!

To clarify some points:

• Regarding file size: I want to confirm - is that 134,217,728 bytes limit for the total size of all files in my blob storage that are indexed by AI search? The PDF I'm trying to index is only around 1000KB (~1MB), which seems well below the limit you mentioned. So I don't think individual file size is the issue.

Content type: PDFs have worked fine previously and are still working for files that were indexed on previous days. This particular issue seems to be only with new files.

Regarding your mention of custom skills: I am using skills in my setup. Here's my current skillset configuration:

• SplitSkill: Splitting document content into pages (5000 char max length)
  • KeyPhraseExtractionSkill: Extracting key phrases from each page
    • OcrSkill: Extracting text from images in the document
      • MergeSkill: Merging OCR text with document content
      About timeout and batch size: Where exactly would I need to increase the timeout from the default 30 seconds to a maximum of 230 seconds? Is this a setting in the indexer configuration or somewhere else? Also, how do I reduce the batch size for my custom skills to ensure they execute within timeout limits? Regarding field mappings: I'm confident this isn't the issue because I'm testing with the exact same document that worked well previously.

One thing I should mention is that I uploaded many files to blob storage at once through an automated crawling process. Could this have caused some kind of queue or processing issue? I've deleted some of these files, but since the retention period is 6 days, they won't be completely removed until then - could this be affecting the indexer?

Would you recommend any adjustments to my skillset configuration or other settings that might help resolve the timeout issue?

Thanks again for your help!

{
  "@odata.etag": "\"0x8DD7D45025A4C72\"",
  "name": "skillset1719988983064",
  "description": "",
  "skills": [
    {
      "@odata.type": "#Microsoft.Skills.Text.SplitSkill",
      "name": "#1",
      "context": "/document/content",
      "defaultLanguageCode": "en",
      "textSplitMode": "pages",
      "maximumPageLength": 5000,
      "pageOverlapLength": 0,
      "maximumPagesToTake": 0,
      "unit": "characters",
      "inputs": [
        {
          "name": "text",
          "source": "/document/content",
          "inputs": []
        }
      ],
      "outputs": [
        {
          "name": "textItems",
          "targetName": "pages"
        }
      ]
    },
    {
      "@odata.type": "#Microsoft.Skills.Text.KeyPhraseExtractionSkill",
      "name": "#2",
      "context": "/document/content/pages/*",
      "defaultLanguageCode": "en",
      "inputs": [
        {
          "name": "text",
          "source": "/document/content/pages/*",
          "inputs": []
        }
      ],
      "outputs": [
        {
          "name": "keyPhrases",
          "targetName": "keyphrases"
        }
      ]
    },
    {
      "@odata.type": "#Microsoft.Skills.Vision.OcrSkill",
      "name": "#3",
      "description": "Extracts text from images",
      "context": "/document/normalized_images/*",
      "textExtractionAlgorithm": "Printed",
      "lineEnding": "Space",
      "defaultLanguageCode": "en",
      "detectOrientation": true,
      "inputs": [
        {
          "name": "image",
          "source": "/document/normalized_images/*",
          "inputs": []
        }
      ],
      "outputs": [
        {
          "name": "text",
          "targetName": "ocr_text"
        }
      ]
    },
    {
      "@odata.type": "#Microsoft.Skills.Text.MergeSkill",
      "name": "#4",
      "description": "Merge OCR text with document content",
      "context": "/document",
      "insertPreTag": " ",
      "insertPostTag": " ",
      "inputs": [
        {
          "name": "text",
          "source": "/document/content",
          "inputs": []
        },
        {
          "name": "itemsToInsert",
          "source": "/document/normalized_images/*/text",
          "inputs": []
        },
        {
          "name": "offsets",
          "source": "/document/normalized_images/*/contentOffset",
          "inputs": []
        }
      ],
      "outputs": [
        {
          "name": "mergedText",
          "targetName": "merged_text"
        }
      ]
    }
  ],
  "cognitiveServices": {
    "@odata.type": "#Microsoft.Azure.Search.DefaultCognitiveServices"
  }
}