0.1.3 - November 11, 2020

• Overall

  • Helm 2.x has reached end-of-life and is no longer supported. Helm 3.x is now required.
  • Support added for the SAS Viya Workload Node Placement
    • By default, monitoring and logging pods are deployed to untainted nodes
    • A new flag, NODE_PLACEMENT_ENABLE supports deploying pods to appropriate
      workload node placement nodes

• Monitoring

  • Several helm charts have moved from stable to prometheus-community.
  • The prometheus-operator helm chart has been deprecated and moved to kube-prometheus-stack.
  • SAS Java and Go ServiceMonitors converted to PodMonitors to properly support merged services

• Logging

  • Support for SAS Viya move to Crunchy Data 4.5
  • Support for changing retention period of log messages
  • Node anti-affinity for Elasticsearch replicas
  • Support for multi-role Elasticsearch nodes (including sample
    to demonstrate usage)
  • Additional documentation on using TLS
  • Removed traces of support for ODFE "demo" security configuration
  • Alternate monitoring solution(proof-of-concept): Fluent Bit ==> Azure Monitor (Log Analytics workspace)