Flow Designer - Internal Server Error while processing paginated Microsoft Graph API response

[uttkarshupa]

Hi everyone,

I'm using a Flow Designer Action to pull devices from the Microsoft Graph API (managedDevices) using @odata.nextLink for pagination. The flow processes records in batches successfully, but after processing around 3,939 out of ~4,752 devices, it consistently fails with an Internal Server Error.

I'm already using pagination (not fetching all records at once), and data is being processed and stored correctly until the failure. Unfortunately, there are no useful logs or stack traces—only the generic Internal Server Error.

Has anyone experienced a similar issue? Could this be related to Flow Designer execution timeout, memory limits, or another platform limitation? Any suggestions on how to troubleshoot or work around this would be greatly appreciated.

Thanks!

[Tanushree Maiti]

Hi @uttkarshupa 

We have flow for fetching data on Azure . Data payload we were sending approx 10K. Same issue we have faced.

We divided the payload in a batch . after that it is working fine. 

You can use also Flow - scheduled job  combination.

[uttkarshupa]

Thanks for your suggestion. We are already fetching the data in batches using Microsoft Graph pagination (@odata.nextLink). Each API call returns around 1000 records, and then we use the next link to fetch the next batch.

The issue is that after processing a few batches (the count varies each run), the flow ends with an Internal Server Error, even though we're not sending all ~5k records in a single request. Did you encounter something similar with Flow Designer timeouts or transaction limits while processing multiple batches in a single execution?

[Ankur Bawiskar]

@uttkarshupa 

why not use Data Stream which handles large incoming request?

Data Stream actions and pagination 

💡 If my response helped, please mark it as correct ✅ and close the thread 🔒— this helps future readers find the solution faster! 🙏

[uttkarshupa]

Thanks for the suggestion. We are currently using a custom Action with Microsoft Graph @odata.nextLink pagination and fetching around 1000 records per request. The flow keeps requesting the next page until all records are processed.

The issue is that the Flow eventually fails with an Internal Server Error after processing several batches, even though each request is limited to ~1000 records.

Would a Data Stream Action help avoid this Flow execution/timeout limitation? Have you used it with Microsoft Graph paginated APIs?