Page Comparison

To be able to handle the KPI management system in a Private Container Deployment, such as Kubernetes, you must prepare a number of scripts according to the instructions below. The scripts that you create are the following:

...

These scripts will be used by different procedures that you find in the sections for KPI management - Distributed Processing.

Preparations before creating scripts:

A Prerequisite is that Spark, ZooKeeper, and Kafka are installed and up and running.

It is mandatory that you need to setup the Apache Kafka host and port as well, run the following commands in your environment:

export KAFKA_BROKERS="127.0.0.1:9092"
export SPARK_UI_PORT=4040 
export MZ_PLATFORM_AUTH="mzadmin:DR-4-1D2E6A059AF8120841E62C87CFDB3FF4"
export MZ_KPI_PROFILE_NAME="kpi_common.SalesModel"
export MZ_PLATFORM_URL="http://127.0.0.1:9036"
export ZOOKEEPER_HOSTS="127.0.0.1:2181"

You also need to export the configuration variables using this command:

export APP_CONFIG="{\"kpi-delay\":\"3\",\"kafka-alarm-topic\": \"kpi-alarm\",\"kafka-input-topic\": \"kpi-input\",\"kafka-output-topic\":\"kpi-output\"}"

Creating scripts:

1. Set up your preferred KPI configuration or use the simplified example configuration, startup the platform. kpi_tst.zip

2. Find and copy the kpi_spark*.mzp among the installation files. Copy it to a place you want to keep your kpi KPI application files.

3. To install the KPI app after building it, and extract the app installation:

...

5. Move the mz_kpiapp folder , to where you prefer to have it and add it to the PATH environment variable.

...