High Query Targeting metric and many inserts

I need to load a large amount of data regularly and need to use update with upsert operation for this due to the fact that data is loaded from several files with one common key.
I create a collection and index by this key, and then use bulkWrite with upsert updateOne, data is read from files and loaded in batches of 500 records with 1 second delay.
As I see, index is used for update operations (by index using statistic from mongo compass), but I don’t understand why Query Targeting metric gives such high results.
Is it normal that in this case the metric produces such large results? And if so, how can this process be organized normally?

1 Like

Hello, we also saw the same issue, although there is a perfectly covered index (scan/get should be 1) the query targeting ratio is actually the same as the operation count. Just want to write cause I want to bump this conversation up.