TimeScale Notes

Installation

https://docs.timescale.com/self-hosted/latest/install/installation-linux

Create Hypertable

CREATE TABLE conditions (
   time        TIMESTAMPTZ       NOT NULL,
   location    TEXT              NOT NULL,
   temperature DOUBLE PRECISION  NULL
);

SELECT create_hypertable('conditions', by_range('time'));
SELECT create_hypertable('conditions', by_range('time', INTERVAL '1 day'));
SELECT set_chunk_time_interval('conditions', INTERVAL '2 days');

The default time interval for a chunk is 7 days.
The dimension can either be by_range or by_hash.

Indexes

You can create indexes on Hyper Tables as per normal.
However when creating a unique index you must include the partioning columns
In the example above that would include the time column.

Compression

Check Distinct Values

-- n_distinct (-1 = Unique, 1 = All The Same)
SELECT tablename, attname, most_common_vals, most_common_freqs, n_distinct 
FROM pg_stats 
-- WHERE tablename = '<TableName>'; 
WHERE schemaname = '<SchemaName>'; 

Compress Table

ALTER TABLE crypto_ticks
SET (timescaledb.compress,
     timescaledb.compress_segmentby='symbol',
     timescaledb.compress_orderby='time DESC');

SELECT compress_chunk(c) from show_chunks('crypto_ticks') c;

Check Compression

SELECT pg_size_pretty(before_compression_total_bytes) as before, 
       pg_size_pretty(after_compression_total_bytes) as after 
  FROM hypertable_compression_stats('crypto_ticks'); 

SELECT pg_size_pretty(hypertable_size('CRYPTO_TICKS'));

Compression Policy

SELECT add_compression_policy('crypto_ticks', INTERVAL '8 days'); 

Links

https://docs.timescale.com/tutorials/latest/financial-tick-data/financial-tick-compress

Since version 2.18, HyperCore has been used for compression.
This uses a hybrid of row and columnar storage.
https://docs.timescale.com/use-timescale/latest/compression (Old)
https://docs.timescale.com/use-timescale/latest/hypercore] (New)