Partition not working in case of Blob Input

Hi M Saad Stream Analytics blob input partitioning currently supports numeric partition values when you configure the partition key in the input settings. String-based folder names are not treated as partitions by the engine, which is why events are not being read when you use strings as partition values.

If you need to organize your data using string-based folder names, you can:

• Use wildcard paths in the input configuration (e.g., .../year=*/month=*/day=*) so that Stream Analytics picks up all matching blobs.
• Add a column inside your event payload to carry the string partition value, and then apply partitioning logic downstream in your query.

For native partition parallelization in Stream Analytics, you will need to use numeric values. String partitions won’t be recognized at input level.

If your scenario depends heavily on string partitions, an alternative is to preprocess or rename blobs (for example, with Data Factory or a storage event trigger) to assign numeric partition keys before feeding them into Stream Analytics.

I hope this information helps. Please do let us know if you have any further queries. Kindly consider upvoting the comment if the information provided is helpful. This can assist other community members in resolving similar issues.

The documentation says that using wildcard (*) is not allowed

https://free.blessedness.top/en-us/azure/stream-analytics/stream-analytics-define-inputs#stream-data-from-blob-storage-or-data-lake-storage-gen2

M Saad Thanks for sharing the doc reference, you’re absolutely right that the * wildcard isn’t supported in the Path pattern field. To clarify my earlier note, Stream Analytics only supports the built-in tokens ({date}, {time}, {partition}) in path patterns, and not the * character.

If your folder structure is string-based, Stream Analytics won’t treat those as partitions at input level. In that case, the recommended options are:

• Include the string value inside the event payload and apply partitioning/filtering logic in your query.
• Preprocess/rename blobs (for example, with ADF or an Event Grid function) into numeric partition folders before ingestion.

That way you’ll stay within the supported pattern syntax and still achieve the logical partitioning you need.

I hope this information helps. Please do let us know if you have any further queries. Kindly consider upvoting the comment if the information provided is helpful. This can assist other community members in resolving similar issues.

M Saad We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.