Forum Discussion

TekkyTec's avatar
TekkyTec
New Member
7 months ago
Solved

Power BI Service Subscription Failure - Intermittent and a single report

Hello Community.  We are having intermittent subscription failures and we have not been able to pinpont why it is occurring.  Wondering if anyone else has had this happen and have a solution to prevent it from reoccurring.

 

Our Power BI service email subscription (snap shot and PDF) for a single small report failed to send on its scheduled daily run. However, the subscription successfully sends when manually triggered. This is the second time this issue has occurred since inception in March 2025 and it appears isolated to this single report subscription.  On the first failure the next day the subscription ran without issue so we left it as is, but this 2nd occurrance causes us angst. We will see if it sends tomorrow without manual intervention.

 

Observed Behavior:

  • Scheduled subscription fails silently without error notifications in the Power BI UI.

  • Manual “send now” successfully sends emails to recipients.

  • Dataset refresh completes successfully prior to scheduled send.

  • No recent changes to report permissions, dataset configuration, or recipient list.

Troubleshooting Steps Taken:

  • Checked dataset refresh history (successful).

  • No service advisories found at the time of issue.

  • Hi TekkyTec 

    This is almost always caused by a temporary capacity or rendering failure at the exact time the scheduled subscription runs, not by the report or recipients. Scheduled subscriptions are queued and executed on Microsoft’s backend rendering service, and if that service is throttled, busy, or briefly unavailable for your capacity at that moment, the job can fail without surfacing a UI error, while a manual “Send now” works because it runs on a different on-demand execution path. That’s why your refresh succeeds, permissions look fine, and it fixes itself the next day. There is no tenant-side fix for this — the only real mitigations are move the report to a less congested capacity (PPU or dedicated Premium), shift the send time away from peak hours, or split the subscription so it doesn’t collide with heavy background activity.

     

3 Replies

  • Power BI subscriptions run in a shared background service that:

    • Renders the report

    • Applies the subscription context

    • Generates the snapshot/PDF

    • Sends the email

    Occasionally, for a specific report, the scheduled job:

    • Times out

    • Fails during background rendering

    • Gets skipped due to transient capacity/service load

    To solve the problem you could:

     

    1) Slightly delay the subscription

    If the subscription runs immediately after refresh:

    • Add a buffer (e.g. 15–30 minutes after refresh completes)

    This avoids race conditions where the model is technically refreshed but not fully available to the subscription renderer.

     

    2) Reduce render complexity for subscriptions

    If possible:

    • Avoid large tables/matrices on the first page

    • Consider a subscription-specific page with fewer visuals

    • Reduce “Show items with no data”

    • Avoid visuals that depend on:

      • SELECTEDVALUE without defaults

      • Dynamic titles driven by slicers

      • Field parameters (these have caused subscription issues in some tenants)

    You don’t need to change the report for users, just make the first page subscription-friendly.

     

    3) Recreate the subscription

    This sounds trivial, but it helps more often than it should:

    • Delete the failing subscription

    • Recreate it from scratch

    • Re-add recipients

    • Re-save

    This can clear corrupted subscription metadata.

  • Hi TekkyTec 

    This is almost always caused by a temporary capacity or rendering failure at the exact time the scheduled subscription runs, not by the report or recipients. Scheduled subscriptions are queued and executed on Microsoft’s backend rendering service, and if that service is throttled, busy, or briefly unavailable for your capacity at that moment, the job can fail without surfacing a UI error, while a manual “Send now” works because it runs on a different on-demand execution path. That’s why your refresh succeeds, permissions look fine, and it fixes itself the next day. There is no tenant-side fix for this — the only real mitigations are move the report to a less congested capacity (PPU or dedicated Premium), shift the send time away from peak hours, or split the subscription so it doesn’t collide with heavy background activity.

     

  • Hi TekkyTec - Below are my suggestions:

    For this scenario:

    1.Delete and recreate the subscription

    2.Move the schedule away from peak times

    3.Consider switching away from PDF if possible

    4. Implement Power Automate for guaranteed delivery

     

    Hope these helps