Activities get retried as their messages no longer exist

[hectorm-bmg]

Hi,

I have a durable function that uses service bus to check on the progress of a process.

The orchestration basically does the following:

• Calls activity "HangRequest".
• "HangRequest" places an item in service bus and comes back to the orchestration
• The orchestration then starts a loop where it calls another activity "CheckStatus"
• "CheckStatus" looks for progress/completion messages from service bus.
• The loop continues until we get a completion message from the service bus.

The problem we are seeing is that the CheckStatus activities are getting rescheduled to be run 5 minutes in the future. The message we are seeing look like this:

Failed to update or delete message [TaskScheduled#1] with ID = c4ceddc9-2288-4e78-a1cc-58b230398d7a because it no longer exists

The orchestration is actually able to complete but these zombie activities keep running for a while.

The HangRequest activity that is outside of the loop does not seem to have this issue. It's only the activities ran in the while loop that present this issue.

[hectorm-bmg]

This issue is a show stopper for our use of durable functions. Any suggestions on how to resolve this would be really appreciated. Thanks!!