Looking at my cluster I clicked the “Performance Advisor” to see if there were any useful indexes I could create. Appeared I have what I need (no slow queries for now). The I selected the tab “Schema Anti-patters”. This displays a potential problem:
Reduce the size of documents
- Average document size is over 2 MB in the collection(s) scanned
However the particular collection has no document in that size range, and the statistic for the colection say:
COLLECTION SIZE: 30.42MB
TOTAL DOCUMENTS: 4838
So I would estimate about 6Kb per document. Which is confirmed when I export a single document as JSON. What could be happening here?
Thanks for any clues,