Its strange. Have you tried agent restart or rather server restart?

Defer save might be happening. Could you write a simple activity to commit it and see that whether that is happening properly? Also, try with commitwitherrorhandeling as well?

may be these test give some answer to this behaviour.

Hi Naveen,

Can you make sure that there are no pending deferred saves that are getting executed after the explicit commit (that is done if any). This could take the instance to have the stale data even though a commit with the new data happens.

Let us know your observation

Regards,

Ratan Balaji.

Have you looked at the audit log?  Anything of interest in it?

Hi Naveen,

I know your facing issues with OOTB agents for debugging purpose you can switch to different modes legacy, standard to observe the behaviour  of the agents.

Thanks,

Arun

Surprisingly and unknowingly issue got resolved. Could not find the exact root cause.

We have same issue in our PROD environment. Case goes to an assignment shape and later SLA gets triggered and flow resumed. The processing continues until the end of the flow - we have multiple connector calls in this part of the flow and have confirmation / data to show that all those connector calls got executed successfully and those external systems updated data on their end too. However, the case came back to same assignment and another agent (at a later point in time) processed it through - (same part of the flow) causing those external calls to be triggered second time. Absolutely no errors in the logs.