Overview

This guide provides the system requirements required for using HCL MaxAI CDM

The system requirements details like:
  • Operating system supported
  • Application server supported
  • Database supported
  • Third-party software required
  • Configurations

Prerequisites

Kubernetes Cluser (K8s) A Kubernetes cluster is required to run Apache Airflow and all Data Engineering (DE) pipeline components. The Bespoke, DBT, and 360 containers (pods) will be automatically deployed within the cluster when the Airflow DAG pipeline is executed.
Unica Platform with system tables running on MariaDB The Unica platform, including modules such as Unica Platform and Unica Campaign, must be installed on a MariaDB database. This MariaDB instance will act as the system database for Unica.
Apache Airflow Version 3.0 or higher Apache Airflow 3.0 or higher is required to orchestrate the DAG (Directed Acyclic Graph) pipeline.

A Persistent Volume (PV) must be configured within Kubernetes to store all DAGs, DDLs (Data Definition Language), configuration files, and other pipeline artifacts

Other Requirements

Network Connectivity from Kubernetes Cluster
Both the UDS Oracle Database and the Unica MariaDB System Database must be accessible from the Kubernetes cluster. All relevant network ports for both databases must be open and reachable from within the cluster to ensure seamless connectivity.
Installation Package
By default, the installation package will be available in the UNICA MaxAI Docker Multiplatform Multilingual e-Assembly in the MHS portal. The installation package includes the following components:
  • Docker image for DBT
  • Docker image for Bespoke
  • Docker image for 360
  • UDS DAG ZIP package, containing:
    • DAG code
    • Database DDLs
    • Pipeline configuration files

Required Python Packages

The necessary Python packages are bundled within the respective Docker images.
  • DBT Docker Image includes:
    • dbt-core==1.9.0
    • dbt-oracle==1.9.1
  • Bespoke Docker Image includes:
    • pandas>=2.2.2
    • numpy>=1.26.4
    • oracledb>=3.1.0
    • psycopg2-binary>=2.9.9
    • PyMySQL>=1.1.1
    • pyodbc>=5.1.0
    • openpyxl==3.1.5
    • requests==2.32

Sizing Recommendations

Recommended resource allocations for key components of the HCL MaxAI Canonical system, including the Airflow orchestrator and the Canonical Database. These guidelines are intended to support the listed data volumes, though actual production environments may require higher specifications based on factors such as concurrency, ETL frequency, and overall data volume.
Table 1. Data Volume Overview
Audiences SRC_CH+SRC_RH
100,000 20 Million
Table 2. Airflow (for DAG Orchestration)
CPU RAM DISK
8 32 GB 500 GB
Note: Actual production environments will vary depending on multiple factors and typically require higher specs depending on concurrency, ETL frequency, query complexity, number of DAGs, task concurrency and data volume. Please considers all such factors while sizing application appropriately.
Table 3. Canonical DB
CPU RAM DISK Oracle Node
16 64 GB 500 GB 1
Note: This includes resources for 5-10 concurrenct users trying to retrieve canonical data via MaxAI.
Table 4. Calculate Disk space for CH/RH tables on RAW + SRC + BDV schemas
Enter Total CH + RH count 1M
Estimated disk space (GB) 0.4
Note: User input required in Enter Total CH + RH count cell.

Accepted values 150K, 3.5M, 1B etc.

Table 5. Calculate Disk space for Audience tables
Enter Audiences count 1M
Estimated disk space for LDZ (GB) 16.78
Estimated disk space for BDV (GB) 19.65
Estimated disk space for ANP (GB) 28.39
Estimated disk space for RDV (GB) 6.08
Estimated disk space for CS_360_DB (GB) 1.29
Total estimated disk space (GB) 72.19
Note: User input required in Enter Audiences count cell.

Accepted values 300K, 2M, 15M etc.

Note: This is minimum sizing. Actual production environments require higher specifications depending on concurrency, ETL frequency, query complexity, number of DAGs, task concurrency, and data volume.

Supported Environment

The Supported Environments section provides a list of required versions for third-party components like the OS, databases, Airflow, and Python.

Supported Third Parties Version
Operating System RHEL 8.x, 9.x
CDM System Database Oracle 19.*
Unica Campaign with System Database MariaDB 10.4.x,10.5.x, 10.6.x
Airflow 3.*
Python 3.10.*