Cassandra Time-Series Cheat Sheet

Last Updated: November 21, 2025

Focus Areas

Focus
`Bucket by day/month`
`Embed TTL per row`


         CREATE TABLE readings (device text, day date, ts timestamp, value double, PRIMARY KEY ((device, day), ts))

Define schema


         INSERT INTO readings (device, day, ts, value) VALUES (...) USING TTL 86400

Write with TTL


         SELECT * FROM readings WHERE device='sensor' AND day='2024-01-01'

Read bucket

Partitioning + TTL keeps Cassandra time-series performant.

💡 Pro Tip: Keep partition rows under 100k for read efficiency.