Optimize your data for real-time analytics

Timescale Cloud: Performance, Scale, Enterprise

Self-hosted products

MST

Hypercore is the TimescaleDB hybrid row-columnar storage engine used by hypertables. Hypertables partition your data in chunks. Chunks stored in the rowstore use a row-oriented data format optimized for high-speed inserts and updates. Chunks stored in the columnstore use a columnar data format optimized for analytics. You ingest hot data into the rowstore. As data cools and becomes more suited for analytics, Timescale Cloud automatically converts these chunks of data to the columnstore. You define the moment when data is converted using a columnstore policy.

When you convert chunks from the rowstore to the columnstore, multiple records are grouped into a single row. The columns of this row hold an array-like structure that stores all the data. For example, data in the following rowstore chunk:

Timestamp	Device ID	Device Type	CPU	Disk IO
12:00:01	A	SSD	70.11	13.4
12:00:01	B	HDD	69.70	20.5
12:00:02	A	SSD	70.12	13.2
12:00:02	B	HDD	69.69	23.4
12:00:03	A	SSD	70.14	13.0
12:00:03	B	HDD	69.70	25.2

Is converted and compressed into arrays in a row in the columnstore:

Timestamp	Device ID	Device Type	CPU	Disk IO
[12:00:01, 12:00:01, 12:00:02, 12:00:02, 12:00:03, 12:00:03]	[A, B, A, B, A, B]	[SSD, HDD, SSD, HDD, SSD, HDD]	[70.11, 69.70, 70.12, 69.69, 70.14, 69.70]	[13.4, 20.5, 13.2, 23.4, 13.0, 25.2]

Because a single row takes up less disk space, you can reduce your chunk size by more than 90%, and can also speed up your queries. This saves on storage costs, and keeps your queries operating at lightning speed.

This page shows you how to get the best results when you set a policy to automatically convert chunks in a hypertable from the rowstore to the columnstore.

Prerequisites

To follow the steps on this page:

Create a target Timescale Cloud service with time-series and analytics enabled.
You need your connection details.

The code samples in this page use the

crypto_sample.zip

data from Try the key Timescale features.

Optimize your data with columnstore policies

The compression ratio and query performance of data in the columnstore is dependent on the order and structure of your data. Rows that change over a dimension should be close to each other. With time-series data, you orderby the time dimension. For example, Timestamp:

Timestamp	Device ID	Device Type	CPU	Disk IO
12:00:01	A	SSD	70.11	13.4

This ensures that records are compressed and accessed in the same order. However, you would always have to access the data using the time dimension, then filter all the rows using other criteria. To make your queries more efficient, you segment your data based on the following:

The way you want to access it. For example, to rapidly access data about a single device, you segmentby the Device ID column. This enables you to run much faster analytical queries on data in the columnstore.
The compression rate you want to achieve. The lower the cardinality of the segmentby column, the better compression results you get.

When Timescale Cloud converts a chunk to the columnstore, TimescaleDB automatically creates a different schema for your data. TimescaleDB creates and uses custom indexes to incorporate the segmentby and orderby parameters when you write to and read from the columnstore.

To set up your hypercore automation:

Connect to your Timescale Cloud service
In Timescale Console open an SQL editor. You can also connect to your service using psql.
Enable columnstore on a hypertable
Create a hypertable for your time-series data using CREATE TABLE. For efficient queries on data in the columnstore, remember to segmentby the column you will use most often to filter your data. For example:
- Use CREATE TABLE for a hypertable
```
CREATE TABLE crypto_ticks (
   "time" TIMESTAMPTZ,
   symbol TEXT,
   price DOUBLE PRECISION,
   day_volume NUMERIC
) WITH (
  tsdb.hypertable,
  tsdb.partition_column='time',
  tsdb.segmentby='symbol', 
  tsdb.orderby='time DESC'
);
```
  If you are self-hosting TimescaleDB v2.19.3 and below, create a PostgreSQL relational table, then convert it using create_hypertable. You then enable hypercore with a call to ALTER TABLE.
- Use ALTER MATERIALIZED VIEW for a continuous aggregate
```
ALTER MATERIALIZED VIEW assets_candlestick_daily set (
   timescaledb.enable_columnstore = true, 
   timescaledb.segmentby = 'symbol' );
```
  Before you say huh, a continuous aggregate is a specialized hypertable.
Add a policy to convert chunks to the columnstore at a specific time interval
Create a columnstore_policy that automatically converts chunks in a hypertable to the columnstore at a specific time interval. For example, convert yesterday's crypto trading data to the columnstore:
```
CALL add_columnstore_policy('crypto_ticks', after => INTERVAL '1d');
```
TimescaleDB is optimized for fast updates on compressed data in the columnstore. To modify data in the columnstore, use standard SQL.
Check the columnstore policy
1. View your data space saving:
  When you convert data to the columnstore, as well as being optimized for analytics, it is compressed by more than 90%. This helps you save on storage costs and keeps your queries operating at lightning speed. To see the amount of space saved:
```
SELECT 
  pg_size_pretty(before_compression_total_bytes) as before,
  pg_size_pretty(after_compression_total_bytes) as after
FROM hypertable_columnstore_stats('crypto_ticks');
```
  You see something like:
  before after
  194 MB 24 MB
2. View the policies that you set or the policies that already exist:
```
SELECT * FROM timescaledb_information.jobs
WHERE proc_name='policy_compression';
```
  See timescaledb_information.jobs.

before	after
194 MB	24 MB

Pause a columnstore policy


SELECT * FROM timescaledb_information.jobs where 
   proc_name = 'policy_compression' AND relname = 'crypto_ticks'

-- Select the JOB_ID from the results
  
SELECT alter_job(JOB_ID, scheduled => false);

See alter_job.

Restart a columnstore policy


SELECT alter_job(JOB_ID, scheduled => true);

See alter_job.

Remove a columnstore policy
```
CALL remove_columnstore_policy('crypto_ticks');
```
See remove_columnstore_policy.
Disable columnstore
If your table has chunks in the columnstore, you have to convert the chunks back to the rowstore before you disable the columnstore.
```
ALTER TABLE crypto_ticks SET (timescaledb.enable_columnstore = false);
```
See alter_table_hypercore.

Timescale Cloud charges are based on the amount of storage you use. You don't pay for fixed storage size, and you don't need to worry about scaling disk size as your data grows—we handle it all for you. To reduce your data costs further, combine hypercore, a data retention policy, and tiered storage.

Reference

For integers, timestamps, and other integer-like types, data is compressed using delta encoding, delta-of-delta, simple-8b, and run-length encoding. For columns with few repeated values, XOR-based and dictionary compression is used. For all other types, dictionary compression is used.

Optimize your data for real-time analytics

Prerequisites

Optimize your data with columnstore policies

Reference

Related Content