Annotation Reference

Annotation reference

Annotation	On	Arguments
@timescale.hypertable	model	column (required), chunkInterval (default "7 days"); partitionColumn + partitions (optional) — a hash space dimension; chunkSkipping (optional) — secondary column(s) to enable chunk skipping on
@timescale.retention	model (also a hypertable)	dropAfter (required) — drop chunks older than this interval
@timescale.compression	model (also a hypertable)	after (required) — compress chunks older than this; segmentBy, orderBy (optional) — columnstore tuning (Prisma field names)
@timescale.continuousAggregate	view	source (a hypertable or another cagg), bucket, timeColumn (required); refresh: { startOffset, endOffset, scheduleInterval }, materializedOnly (optional)
@timescale.bucket	view field	— (exactly one per aggregate)
@timescale.groupBy	view field	—
@timescale.aggregate	view field	fn (avg|sum|min|max|count), column

See: Hypertables, Continuous aggregates, Retention & compression.

Intervals

Intervals are "<amount> <unit>" where unit is any PostgreSQL interval input unit — microsecond(s), millisecond(s), second(s), minute(s), hour(s), day(s), week(s), month(s), year(s), decade(s), or centur(y|ies) — validated at compile time and runtime ("1 hour", "7 days", "2 years").

quarter is not an interval unit in PostgreSQL; combined forms like "1 year 2 months" aren't supported by this single-unit type.

Space partitioning

Add partitionColumn + partitions to @timescale.hypertable for a hash space dimension (add_dimension(..., by_hash(column, partitions))) on top of the time dimension:

/// @timescale.hypertable(column: "time", chunkInterval: "1 day", partitionColumn: "deviceId", partitions: 4)
model SensorReading {
  time     DateTime
  deviceId Int

  @@id([deviceId, time]) // a partitioning column must be part of the table's PK / unique key
}

partitionColumn takes a Prisma field name (mapped to its @map column); partitions is a positive integer. It's transparent to timeBucket queries and survives prisma migrate reset. More in Hypertables.

Chunk skipping

Add chunkSkipping (a Prisma field name, or several comma-separated) to @timescale.hypertable to record per-chunk min/max range stats on a secondary column so the planner can exclude non-matching compressed chunks. Remember to turn timescaledb.enable_chunk_skipping on at query time — see Hypertables → Chunk skipping for the full story and caveats.