Installing and Setting Up ClickHouse

ClickHouse is an open-code columnar DBMS that allows for executing analytical queries in real time based on structured big data.

NOTE. If ClickHouse is set up incorrectly, it may result in data model work restrictions.

Step 1. Installing ClickHouse

To install ClickHouse:

  1. Create the clickhouse-odbc driver manually based on the resources corresponding to the used version of Linux OS and install the driver to a computer with installed BI server.

  2. Install the clickhouse-client driver to a computer with installed BI server.

  3. Grant folder access permissions:

    • /var/lib/clickhouse/user_files:

sudo chown www-data:www-data /var/lib/clickhouse/user_files

sudo chmod 755 /var/lib/clickhouse

  1. Restart clickhouse-server:

sudo systemctl restart clickhouse-server

  1. Install the clickhouse-odbc driver to a computer with installed BI server.

  2. Install the clickhouse-client driver to a computer with installed BI server using one of the methods:

    • In the Docker virtual environment start the clickhouse container:

docker run -d --name clickhouse --volume="<path to folder with temporary files>" --ulimit nofile=262144:262144 clickhouse/clickhouse-server

In the <path to folder with temporary files> specify path to folder with temporary files inside the container. For example: temp:/tmp.

Step 2. Setting Up ClickHouse

After installing the clickhouse-odbc and clickhouse-client drivers determine the following:

  1. Create the Database object in the repository with connection to the ClickHouse server. A detailed description is given in the Connecting to Relational Databases section.

  2. Create the repository variable EXTRACT_DB. As a value, specify ID of the Database object.

  3. Create a service user in the repository who will have access to required objects (models), set the Login, Login to Object Navigator and Read and Open All Objects privileges.

  4. Start the PP.Util utility under the www-data user in Astra Linux, Ubuntu or the apache user in CentOS, RED OS, Rocky Linux, ALT Linux. Next, save service user credentials. The example of command:

sudo -u www-data PP_NODISPLAY=1 /opt/foresight/fp10.x-biserver/bin/PP.Util /save_creds mb_defenition /DEC login pass

Where:

After executing the command service user credentials are saved to the Metabases.xml.

In case of error

See also:

Building Data Model