> “...But this is not the case! Some of these parts have never been the case - MongoDB has never been "eventually consistent" - updates are streamed and applied sequentially to secondary nodes, so although they might be behind, they're never inconsistent.”<p>Isn’t this exactly what eventual consistency is? Or am I missing something? If a client performs a read query on the primary, followed by a second read on a secondary (perhaps because the primary server is unavailable), then it seems clear that there could be a chance of reading stale data.