Dear experts,
Scenario:
Message is sent from SWMS to ECC via PI.
message originally failed in Integration Engine due to some system issue(not remembering clearly the accurate code/descriptions), seems ECC was under maintenance at that time.
Then the message was re-processed by someone after ECC was up, and failed with a different error information -- mapping error. Here comes my another question, ideally the message should be sent out successfully this time, how the same message failed with two different reason?
The message sent from SWMS was correct, no data issue with it as it can be processed successfully in Message Mapping, then we noticed a interesting thing in IE:
when we double clicked this message in sxmb_moni, the pipeline steps got displayed on the left side of the window, ideally the sequence should be <validation> -> receiver determination -> interface determination -> request message mapping -> <validation> -> call adapter.
However, the sequence was below:
receiver determination -> interface determination -> reprocess(should be caused by restart..) -> call adapter -> request message mapping ...
here the mapping was requested after call adapter, and we assume this was the reason it failed with mapping error, as it is using the output to do the mapping again, and we cancelled this message and asked SWMS to resent the message.. Q1: Not sure whether this assumption is correct or not, if correct, how can it happen in this way.
Q2: (more important question): Later the ECC guys confirmed they got the message which we cancelled above -- when they searching the message ID it was shown successfully in ECC sxmb_moni... How a cancelled message got sent to receiver.? system display error?
Any explanation would be greatly appreciated, thanks !
(As I am writing this issue at home without access to PI system, some description may not be exactly the same or not complete, please kindly point out if you want any more information, thanks! )