This file attempts to provide information useful to mailbox/cassandra developers.
Can be edited using this source file on this website.
This file should be edited everytime the table structure is modified.
The Cassandra mailbox is heavily relying on denormalization. The same data is stored several times, in distinct tables, to satisfy the various read access patterns. Of course, the various tables can become inconsistent between them.
We have been developing several approaches in order to diagnose, mitigate and solve such issues. They are described in the following documents:
