If the user has provided Elasticsearch configuration in the FEPLogging CR, and FEPCluster is configured to send server log files and auditlog files to that FEPLogging instance, those logs will be visible on Elasticsearch stack or Elastic Cloud. Assuming Elasticsearch has been configured with Kibana then logs will be visible in Kibana Dashborad. User can use fep log csv fields to create various Dashbord in Kiabana as well. LogstashPrefix and auditLogstashPrefix will be used to filter logs of specific FEPLogging instance.

User can verify if FEPLogging feature is configured properly or not by checking real time FEP logs are populating to the destination.

FEPLogging feature is used for raising alarm/alert based on postgres severity counts as well. While user creates FEPLogging CR, Operator will forward real time counts of various postgres seviriry metrics to Openshift managed Prometheus. Openshift managed Alertmanager can access this metrics counters and user can use them to create alerts/alarms. There are 4 default alert rules already created as part of FEPLogging implementation as listed below:

• FEPLogErrorMessage

• FEPLogFatalMessage

• FEPLogPanicMessage

• FEPLogWarningMessage

Prometheus will scrape postgres_severity counter at every 30s as default scrape interval is 30s. User can modify this scrape interval from FEPLogging CR. After each scrape interval, if any change/increment found in postgres_severity counter then alert rule will be fired. User can check counts of postgres_severity metrics anytime from Prometheus dashboard as well.

In order to forward auditlog to Elasticsearch, update the FEPCluster to enable creating auditlog.

Example)

spec:
  fep:
    fepChildCrVal:
      customPgAudit: |
        [output]
        logger = 'auditlog'
        log_directory = '/database/log/audit'
      customPgParams: |
        shared_preload_libraries='...,pgaudit'
        session_preload_libraries='...,pgaudit'