Build fluency in the vocabulary of a database purpose-built for storing timestamped data.
0 / 5 completed
1 / 5
At standup, a dev mentions a database purpose-built for storing and querying timestamped data points, with its storage and compression specifically optimized around the time dimension. What is this kind of database called?
A time-series database, or TSDB, is purpose-built for storing and querying timestamped data points, with storage and compression specifically optimized around the time dimension, unlike a general-purpose relational database with no such specialization. A general-purpose database handles a timestamped record adequately at small scale, but its storage and query patterns aren't tuned for the volume and shape of high-frequency time-series data. This specialization is what lets a TSDB handle a much higher-volume metrics workload efficiently.
2 / 5
During a design review, the team wants older, fine-grained data points pre-aggregated into a coarser resolution automatically, rather than keeping every raw data point indefinitely at full resolution. Which capability supports this?
Downsampling, or rollup, automatically pre-aggregates older, fine-grained data points into a coarser resolution, rather than keeping every raw data point indefinitely at full resolution. Keeping every raw point forever at full resolution causes storage to grow without bound as time series accumulate. This downsampling is what keeps a time-series database's storage growth manageable while still preserving a useful, if coarser, view of older historical data.
3 / 5
In a code review, a dev notices a retention policy automatically expires and deletes data older than a configured window, rather than requiring a person to manually clean up old data periodically. What does this represent?
An automatic retention policy expires and deletes data older than a configured window without requiring a person to manually clean up old data periodically. Requiring manual cleanup risks that task being forgotten or delayed, letting storage grow unnecessarily in the meantime. This automatic expiration is what keeps a time-series database's storage bounded to only the retention window that's actually needed for the workload.
4 / 5
An incident report shows a time-series database filled up its disk entirely because raw-resolution data had been retained indefinitely with no downsampling or retention policy ever configured. What practice would prevent this?
Configuring downsampling for older data, alongside a retention policy that expires data past a defined age, keeps storage growth bounded rather than accumulating every raw data point forever. Retaining everything indefinitely with no such configuration is exactly what causes the disk to eventually fill up, as this incident describes. This combination of downsampling and retention is standard, essential configuration for any production time-series database.
5 / 5
During a PR review, a teammate asks why the team uses a purpose-built time-series database for high-volume metrics instead of a general-purpose relational database. What is the reasoning?
A TSDB's storage, compression, downsampling, and retention are all optimized around the time dimension, handling a high-volume, timestamped workload far more efficiently than a general-purpose row store not designed for that specific access pattern. A general-purpose database can technically store a timestamped record, but it lacks these built-in optimizations at scale. The tradeoff is the added operational complexity of running and learning a separate, specialized database system alongside a team's existing general-purpose one.