We have a few workflows that run quite a bit of data through them (over 5 mill records). We are splitting out by recency in our database and have multiple deliveries (with sometimes AB testing).
Once in a while we get error message:
Status: Personalization or Count Failed
The delivery has failed
XSV-350122 An error occurred and the process has been stopped. SOP-330011 Error while executing the method 'PrepareMessage' of service 'nms:delivery'. SOP-330011 Error while executing the method 'PrepareMessageImpl' of service 'nms:delivery'. PGS-220000 PostgreSQL error: ERROR: duplicate key value violates unique constraint "nmsbroadlogrcp_id" DETAIL: Key (ibroadlogid)=(2062510602) already exists.
We have added in dedupe on the workflow, added in waits between firing off next delivery, waves for processing and put the delivery on low urgency. The workflow processes fine and it always seems to be the individual deliveries that have the problem.
I don't really know what more to do to stop this from happening. I should point out that using the same data set this doesn't happen all the time. Maybe 1/20 messages we send to effectively the same data set with the same conditions - but potentially adding in more deliveries. Is it purely the number of deliveries in the workflow?