Learning Apache Cassandra

Reversing the order of rows

We're on our way to building great data access logic for displaying a user's timeline of status updates. There's one minor problem though; if I'm viewing alice's status updates, I'm probably interested in reading her most recent ones first. So far, we've always gotten the status updates in ascending order of id, meaning the oldest ones come first. Fortunately, we've got a couple of ways to change this.

Reversing clustering order at query time

Using our existing user_status_updates table, we can instruct Cassandra to return results in reverse order of id:

SELECT "id", DATEOF("id"), "body"
FROM "user_status_updates"
WHERE "username" = 'alice'
ORDER BY "id" DESC;

This is the first time we've seen an ORDER BY in CQL, but it ...