Configuring installation files

Once you have the prerequisites in place, you can deploy the OMEGAMON® AI Insights Docker application. You need to configure the installation files.

1. Configure the .env_prod file located in the your_kmua_directory/install/ directory.

   This file contains the main customization parameters used by Docker.

   # Python Services 
   HOST_ADDRESS=<hostname> <- update your kmua server name, ex: server.company.com 
   
   ## SSL JKS 
   JKS_TRUSTSORE=<path-to-truststore.jks> <- update your truststore path, ex: /etc/ssl/server/truststore.jks 
   
   ## ML Shared folders 
   ML_CONFIG=<path-to-file-settings.toml> <- path and name of the ML config file, ex: ./config-python/settings.toml        
   ML_STORE_FOLDER=<path-to-models-folder> <- path of the model storage directory, ex: ./models   
   CONFIG_SERVER_FOLDER=<path-to-config-java-folder> <- path of the services config, ex: ./config-java   
   
   ## Logs 
   LOG_FOLDER=<path-to-log-folder> <- root path of the log files for all containers 
   LOG_DOCKER_SIZE=<docker-max-size-log> <- max size of individual log file 
   
   ## Timezone 
   TIMEZONE=<timezone> <- timezone of the containers, should be synchronized with host, ex: Europe/Zurich

2. Configure the ml-orchestrator.yml file located in the your_kmua_directory/install/config-java/ directory.

   This file contains the customization parameters used by the Java part of OMEGAMON AI Insights. This configuration file is divided into three parts:

   Elasticsearch

   # TLS Config. 
   
   ssl: 
     truststorePath: <path-to-truststore.jks>     <- update your truststore path + filename 
     truststorePassword: <password-for-truststore.jks>    <- update your truststore password 
   
   # Elasticsearch Config. 
   elasticsearch: 
     protocol: <elasticsearch-protocol> 	<- Elasticsearch communication protocol 
     host: <elasticsearch-hostname> 	<- Elasticsearch server hostname 
     port: <elasticsearch-port> 		<- Elasticsearch port number 
     username: <elasticsearch-username> 	<- Elasticsearch username 
     password: <elasticsearch-password> 	<- Elasticsearch password 
   

   Kibana reference, alerting parameters

   You can modify each context alert mail content in this section. For more information, see Alerting.

   # Kibana Config. 
   kibana_scheme: <kibana-protocol>  		<- Kibana communication protocol 
   kibana_host: <kibana-hostname> 		<- Kibana server hostname 
   kibana_port: <kibana-port> 			<- Kibana port number 
   kibana_username: <kibana-username> 		<- Kibana username 
   kibana_password: <kibana-password> 		<- Kibana password 
   
   # SMTP Config. 
   spring: 
     mail: 
       host: <mailserver-host> <- smtp hostname 
       port: <mailserver-port> <- smtp port 
       username: <mailserver-username> <- smtp credential 
       password: <mailserver-password> <- smtp password 
   
   # Mail 
   sender_mail: sender@company.com  <- sender's mail address 
   receiver_mail: customer@company.com <- receiver's mail address 
   customer_name: <customer-name> <- name of customer 
   sender_support_contact: <support-contact> <- sender's support contact 
   sender_name: <sender-name> <- sender name 
   sender_designation: <sender-designation> <- sender designation 
   sender_company_name: <sender-company-name> <- sender's company name 
   sender_contact_details: <sender-contact-details> <- sender's contact details 

   Scheduling parameters

   Scheduling parameters use spring crontab syntax. For more information, refer to specific documentation.

   This set of parameters is an example of configuration, including all contexts. You must adapt it to your installation by keeping only the contexts used.

   # Scheduler Tasks Config. 
   scheduler:  
     alert: 
       enabled: true  <- enable alerting  
     predict: 
       enabled: true  <- enable forecast processing, do not change 
     train: 
       enabled: true  <- enable training processing, do not change   
   contexts: 
       #---------------------------------KZHS--------------------------- 
       - name: kmua-zos-hourly-sysplex 
         alert_cron: 0 12 * ? * * 
         predict_cron: 0 2 * ? * *      # each hour at 2 min 
         training_cron: 0 7 0 ? * 0,3   # Sunday,Wednesday at midnight at 7 min 
       #---------------------------------KJHJ--------------------------- 
       - name: kmua-jvm-hourly-job 
         alert_cron: 0 16 * ? * * 
         predict_cron: 0 6 * ? * *      # each hour at 6 min 
         training_cron: 0 7 0 ? * 2,5   # Tuesday Friday at midnight at 7 min 
       #---------------------------------KNHL--------------------------- 
       - name: kmua-network-hourly-lpar 
         alert_cron: 0 14 * ? * * 
         predict_cron: 0 4 * ? * *       # each hour at 4 min 
         training_cron: 0 7 0 ? * 1,4    # Monday Thursday at midnight at 7 min 
       #---------------------------------KCHRC--------------------------- 
       - name: kmua-cics-hourly-region-cpu 
         alert_cron: 0 18 * ? * * 
         predict_cron: 0 8 * ? * *       # each hour at 8 min 
         training_cron: 0 37 0 ? * 3,6   # Wednesday Saturday at midnight at 37 min 
       #---------------------------------KCHRR--------------------------- 
       - name: kmua-cics-hourly-region-rt 
         alert_cron: 0 20 * ? * * 
    predict_cron: 0 10 * ? * *      # each hour at 10 min 
         training_cron: 0 37 0 ? * 1,4   # Monday Thursday at midnight at 37 min 
       #---------------------------------KZDL--------------------------- 
       - name: kmua-zos-daily-lpar 
         alert_cron: 0 30 3 ? * * 
         predict_cron: 0 0 1 ? * *            # each day at 1 am 
         training_cron: 0 37 1 13,26 * *      # each 13 and 26 of each month at 1:37 am 
     #---------------------------------------------------------------- 

   Scheduling tasks

   Efficiently automating specific tasks or processes can be achieved by scheduling them to run at predefined intervals, times, or triggers. This practice ensures optimal resource utilization and timely execution. The scheduling frequency might vary based on the context, and you have the flexibility to determine how frequently tasks should run – be it hourly, daily, weekly, monthly, or customized intervals. OMEGAMON AI Insights triggers these tasks according to the scheduled time.

   To facilitate a well-organized and responsive monitoring system tailored to your specific needs, adhere to the following guidelines. Each context has distinct scheduling requirements:

   Table 1. Scheduling requirements of contexts

   Context	Predict	Train	Alert
   kmua-zos-hourly-sysplex	Each hour	Twice a week	Each hour after predict
   kmua-zos-daily-lpar	Each day	Twice a month	Each day after predict
   kmua-network-hourly-lpar	Each hour	Twice a week	Each hour after predict
   kmua-jvm-hourly-job	Each hour	Twice a week	Each 4 hours
   kmua-cics-hourly-region-cpu	Each hour	Twice a week	Each hour after predict
   kmua-cics-hourly-region-rt	Each hour	Twice a week	Each hour after predict

3. Configure the settings.toml file located in the your_kmua_directory/install/config-python/ directory.

   This file contains the customization parameters used by Python part of OMEGAMON AI Insights:

   [default] 
   elasticsearch = { host = "<elasticsearch-hostname>", port = <elasticsearch-port>, username = "<elasticsearch-username>", password = "<elasticsearch-password>", scheme = "<elasticsearch-protocol>", path_to_elastic_certs = "./ca.pem", verify_certs = true } 
   
   [kmua-cics-hourly-region-rt] 
   iqr_settings = { transaction_col = "transactions_total", activity_col = "activity_level", transaction_threshold = 600, ratio_threshold = 0.01 } 
   

   You need to update the Elasticsearch parameters for enabling the communication between the Machine Learning module and Elasticsearch server.

   Note: The second configuration line is dedicated to CICS Response Time Use Case. It allows you to customize thresholds for adjusting active and inactive regions. You need to contact the Rocket support for advice about these parameters.