Unica Journey Hardware Sizing Details

This hardware sizing guide provide information on the hardware requirements for installing Unica Journey. The calculations mentioned are based on assumptions related to application configuration and tuning. The guidelines shared are for information purpose only.

Disclaimer - This guide illustrates the varying hardware sizing requirement, based on various assumptions. This guide does not replace the need for specific Infrastructure details before the deployment. The mentioned consumptions are based on a specific scenario and hardware. Proper estimation should be done before buying the hardware. On top of the suggested hardware, appropriate buffer should be added for any expected peak/future load.

This document provides general guidelines for Journey deployment at your on-premise data centre or virtualized cloud environment.

Journey Sizing Guidelines

  • Define Marketing Objectives
  • Define Marketing Use Cases
  • Define Marketing Processes
  • Define IT environments & IT Integrations
  • Implementation Planning & Execution
  • Steady-State Operations – Journey design, execution & performance reviews
  • Periodic Reviews of marketing campaigns, processes & outcomes

Recommended Journey Components Deployment

You can refer to Journey Installation guide for information related to deployment of Unica products with Unica Journey. Journey includes different components which can be deployed on same or distributed servers. It is recommended to install all Journey components on different server for optimized performance. This guide can be used for sizing exercise for below Journey components.
  • Journey - Web
  • Journey - Engine
  • Kafka Server
  • Kafka (Zookeeper)

Factors to Consider in Journey Sizing

  • Number of simultaneous Journeys
  • Average Audiences per Journey
  • Average number of Control Points in each Journey
  • Average Delay in each Journey
  • Average number of Touch Points – Email, SMS, CRM – that require external integrations
  • Number of Data Definition fields containing Journey data.

Journey Complexity

This sizing recommendations are based on Journey complexity. Following factors to be considered for performance calculations.
  1. Small Journey
    Based on the usage of below components we can size Journey as Small:
    • Number of audiences: Upto 20
    • Number of control touch points: Upto 20
    • Number of delay touch points: Upto 10
    • Number of decision split: Upto 10
  2. Medium Journey
    Based on the usage of below components we can size Journey as Medium:
    • Number of audiences: 20-40
    • Number of control touch points: 20 - 40
    • Number of delay touch points: 10-20
    • Number of decision split: 10-20
  3. Large Journey
    Based on the usage of below components we can size Journey as Large:
    • Number of audiences: 40-60
    • Number of control touch points: 40-60
    • Number of delay touch points: 20-30
    • Number of decision split: 20-30

Calculating Journey Sizing Requirement

Based on the above Journey complexity, you can calculate the Journey environment hardware requirements using below link: Will add link here:

Disclaimer

The above does not consider any backup and archiving needs. Persistent storage - file and database - shall increase rapidly. Any available hardware cache shall improve performace.

Recommended configuration for replication of kafka messages

Based on the number of Kafka brokers you are planning to deploy in your environment, you will need to update the replication factors in Journey configuration:
Number of Kafka Broker Replication factor
1 1
2 1
3 2
5 3
7 5
Note: Property offsets.topic.replication.factor governs the value of kafka replication.
No of partitions of kafka topic should always be equal to max number of engine service instances.
For enabling Kafka security, required configuration in below file
  1. Journey web : application.properties
  2. Journey engine : journey_master_config.properties & log4j2.xml

Journey Engine Configuration

Data Size OS CPU Default Instances of Engine Services Data Batch size(Recommended)
1,00,000 Windows 6

Data Clean -2

Data Batch -2

Data Map -2

*Explicit Service default max instance should be 1"

 databatch.timeout =10
Linux 4 databatch.batchSize =100
5,00,000 Windows 6

Data Clean -4

Data Batch -3

Data Map -3

*Explicit Service default max instances should be 2

 databatch.timeout =10
Linux 4 databatch.batchSize =100
10,00,000 Windows 12

Data Clean -5

Data Batch -5

Data Map -5

*Explicit Service default max instances should be 5

 databatch.timeout =15
Linux 8 databatch.batchSize =200
Note: Following three properties topic.partitions, ​topic.max.instance and ​default.instance controls the Journey instances. A service cannot have max instances more than topic. partitions and default.instance cannot go less than 1.

Guidelines for Changing Assumptions

Journey is a database-centric application and performance of database server is very critical. Running workflows, segmentation, tracking data uploads, inbound Interactions, analytics and other activities all generate database activity. The size and frequency of these operations determine the size of your database servers. Assumptions illustrated in the senarios have significant impact on the hardware requirement and deployment architecture. This section discusses about the various guidelines for different assumptions.