Hi,
This is a tricky scenario.I tried this out by going directly to the delete program (RPCIPDEL) but it throws the same error.
The next logical step would be to try to reverse the posting and then re-post it with the connection restored and see if that works. Did you try this out?
I went ahead and debugged the logic directly from PCP0 and found that the standard program throws the error out within Program SAPMPCP0 and INCLUDE MPCP0F01 within subroutine delete_run. Here is a screenshot of the code that actually throws the error out.
If reversing the posting and reposting the run once the connection is restored does not work then this is going to be difficult. The last ditch effort would be to login with firefighter access, set a breakpoint and trick the system into thinking it is in a different status by changing the variable in the debugger in which case it should go into the delete logic within RPCIPDEL.
Theoretically this should delete everything as it would if it were in a different status, but in a production environment theoretically usually isnt good enough. You may need to test this out in the QA system by deleting something that is in a status 50 and make sure it does not screw anything else up (PCALAC, Payroll results etc).
My thought is that deleting a run that should not be deleted would screw things up on the FI side more than HCM and if that is the case then you might be alright as long as nothing was transferred to FI at all which appears to be the case from you saying the connection was not there.
Thanks,
Imran