Using PostgreSQL and TimescaleDB with Ignition
Install PostgreSQL
- Download Postgres
- Run the installer.
Install TimescaleDB
TimescaleDB is an extension for PostgreSQL database. Follow the installation guide below for getting the extension installed on your computer.
Installation Guide and Download
-
Download TimescaleDB installer zip file and extract the
timescaledb
folder to the desktop. -
Add Postgres file path to system environment variables.
-
Search Environment Variables and click
Edit the system environment variables
-
When the system properties windows comes up, make sure it's on the
Advanced
tab and click theEnvironment Variables
button. -
Click on the
Path
variable in the System variables table and click theEdit...
button. -
Double-click the next empty row in the table and paste in the path to the bin folder in the PostgreSQL installation folder.
Example path:
C:\Program Files\PostgreSQL\17\bin
-
Click
OK
. - Try running
pg_config
from a command line to confirm the path is working.
-
-
Stop the PostgreSQL service.
net stop postgresql-x64-17
-
Right click on
setup.exe
within the extracted TimescaleDB folder and click Run as administrator. A command prompt will appear. -
Press
n
to skip tuning the PostgresDB installation -
Open a command prompt inside the installer folder and at the command prompt, run the following command using the appropriate PostgreSQL data path and hit
enter
Example command line to append configuration to the end of the exising configuration:
timescaledb-tune --quiet --yes --dry-run >> "C:\Program Files\PostgreSQL\17\data"
Example command line to update the PostgreSQL configuration:
timescaledb-tune --quiet --yes "--conf-path=C:\Program Files\PostgreSQL\17\data"
- If the installation is not successful due to an "access denied" error, make sure you ran the setup.exe as administrator.
- Start the PostgreSQL service:
net start postgresql-x64-17
Setup PostgreSQL for Ignition
Configure PostgreSQL for remote connections
Edit the following file with a text editor:
C:\Program Files\PostgreSQL\17\data\pg_hba.conf
Change the lines that look like this:
# IPv4 local connections:
host all all 127.0.0.1/32 scram-sha-256
to this:this (adjust the additional line to a specific subnet if you want to restrict access to specific IPs rather than allowing any IPv4 connection):
# IPv4 local connections:
host all all 127.0.0.1/32 scram-sha-256
host all all 0.0.0.0/0 scram-sha-256
Creating the Ignition user/login
CREATE ROLE ignition WITH
LOGIN
SUPERUSER
CREATEDB
CREATEROLE
INHERIT
REPLICATION
BYPASSRLS
CONNECTION LIMIT -1
PASSWORD 'ignition_password';
Create the Historian, AlarmLog, and AuditLog databases
CREATE DATABASE historian
WITH
OWNER = ignition
ENCODING = 'UTF8'
LOCALE_PROVIDER = 'libc'
CONNECTION LIMIT = -1
IS_TEMPLATE = False;
CREATE DATABASE alarmlog
WITH
OWNER = ignition
ENCODING = 'UTF8'
LOCALE_PROVIDER = 'libc'
CONNECTION LIMIT = -1
IS_TEMPLATE = False;
CREATE DATABASE auditlog
WITH
OWNER = ignition
ENCODING = 'UTF8'
LOCALE_PROVIDER = 'libc'
CONNECTION LIMIT = -1
IS_TEMPLATE = False;
Configure Ignition
Set up 3 databases (one each) for the previous databases using the correct IP, database name, and login credentials. These should be named:
- AlarmLog
- AuditLog
- Historian
Now go to Alarming -> Journal and create a new database Alarm Journal Profile and name it AlarmLog and point it to the AlarmLog datasource. Enable Data Pruning and set the Pruning Age appropriately.
Now go to Security -> Auditing and create a new database Audit Profile and name it AuditLog and point it to the AuditLog datasource. Enable Data Pruning and set the Pruning Age appropriately.
Setup TimescaleDB for Ignition
Now we need to go into Postgres and setup TimescaleDB for use with the Ignition tables. Use a database editor to run the following SQL commands. pgAdmin comes with Postgres and can be used.
Create a database for Ignition to use, or use your existing Ignition Postgres database. Use this database when performing the following commands.
Load the extension into the database.
CREATE EXTENSION IF NOT EXISTS timescaledb;
Create the hyper-table
This will partition the table into 24 hour chunks.
SELECT * FROM create_hypertable('sqlth_1_data','t_stamp', if_not_exists => True, chunk_time_interval => 86400000, migrate_data => True);
Set up the hyper-table
Add compression to the table.
ALTER TABLE sqlth_1_data SET (timescaledb.compress, timescaledb.compress_orderby = 't_stamp DESC', timescaledb.compress_segmentby = 'tagid');
Create a function that will help translate your time stamp columns format for TimescaleDB.
CREATE OR REPLACE FUNCTION unix_now() returns BIGINT LANGUAGE SQL STABLE as $$ SELECT (extract(epoch from now())*1000)::bigint $$;
Set integer now function
SELECT set_integer_now_func('sqlth_1_data', 'unix_now');
Add compression policy to compress chunks older than 7 days
SELECT add_compression_policy('sqlth_1_data', 604800000);
Run the following command to automatically drop chunks older than the specified cutoff.
1 Year
SELECT add_retention_policy('sqlth_1_data', 31536000000);
2 Years
SELECT add_retention_policy('sqlth_1_data', 3072000000);
5 Years
SELECT add_retention_policy('sqlth_1_data', 157680000000);
10 Years
SELECT add_retention_policy('sqlth_1_data', 315360000000);
Otherwise, you must manually run the following query on a schedule to re-create the same functionality using values from above (1 year shown).
SELECT drop_chunks('sqlth_1_data', 31536000000);