Someone in charge of customer journeys discovered duplicates in a resulting dataset based on one of our data extensions; we are now tasked with routing out the source of said duplicates.
The odd thing about it is that these duplicates exist despite a three part primary key across the entries. Below is included examples of the data extension in question and the violating entries within them.
This did leave us quite perplexed as our earlier experiments showed us that we were not able to violate the primary keys in this way.
Therefore we took a ‘copy’ of the dataextention in question, and tried to populate it with duplicate data, we were allowed to do that until we reached about 3 entries then we would have to delete some to insert more duplicates. However this behavior ceased to happen around 09:00 am (EST) yesterday.
Now whenever I try to upsert the involved data entries I am met with either a response 200 from the REST api, I assume it updates, or a generic error message in the sales force marketing cloud UI.
The duplicate entries still exist though, I am just unable to replicate the behavior that created them at the moment.
Questions are as follows.
- Do Primary keys work as one would assume? (unique not null, more than one makes it composite
- How did I insert the duplicate values?
- Have anyone else experience something similar?
I hope that someone within the community is able to shed some light on these events.
Minor note: i was not able to replicate the duplicating experience if i manually created a new data extension with an identical setup only, if i did a 'copy',does this mean that it is possible to 'corrupt' a data extension?