Change History of Kubernetes Data Collector

Amended limit/request sizings to decrease CPU and memory footprint for size0 and size1 configurations

Added four new monitored types: (1) Openshift Route, (2) Openshift Cluster Resource Quota, (3) Kubernetes Resource Quota, and (4) Openshift DeploymentConfig (added under the Kubernetes Deployment entity type)

Improvements to Kubernetes Namespace/Openshift Project dashboard. New widgets include aggregate metrics for CPU, memory, network traffic (transmitted/received), and network errors transmitted/received) and 'Top 5 Pods' widgets for CPU and memory.

Propagation of events to Kubernetes Namespace/Openshift Project timeline (via inclusion of namespace in event payload's 'relatedResources'). Pod 'Normal' events (such as created container/pulled image) will be filtered from this view to prevent noise.

New widgets and dashboards for Openshift Cluster Resource Quota, Kubernetes Resource Quota, and Openshift Route

Simplification of Kubernetes Service thresholding

Ability to set thresholds on Kubernetes Deployment, Daemon Set, Job, Replica Set, Replication Controller, and Stateful Set

Remove 'Normal' CronJob events from cluster timeline, greatly reducing noise

Reduce outgoing requests by metric-collector container by 90x, greatly reducing CPU usage of container

Extended multiprocessing and session management in the metric collector to decrease total runtime for collection and submission 

MCM Subscription subgraph enrichment to support RBAC

Support for Power and zLinux

Amended display name columns for resource lists to be more consistent with kubectl/oc console

Extended detail in 'Details' event payload

Special character '-' now supported for 'clusterName' and changing 'clusterName' post-deployment is now not recommended

Specify PYTHONUNBUFFERED=1 to prevent logs from hanging

Added descriptions for Kubernetes/Openshift top level metadata objects

General bug fixes and improvements

Collection/submission of events sourced from all entity types

Performance enhancements for cache initialization

Collection and linking of MCM subscriptions to Kubernetes controllers (deployments, statefulsets, etc.)

Change CEM event correlation field from 'application' to 'controller'

Denote sender.component field as 'icam' (backend performance enhancement)

Amend Golden Signals default tabs

Add metadata constructs to enable setting thresholds from RUM object contexts

Add metadata constructs to enable baselining from Kubernetes Service

Amend relationship direction and type from k8sNode -> k8sNamespace and k8sNamespace -> others

SRE 'golden signals' on service dashboard

Extended 'Display Columns' in list views

CPU Graph Units (report in millicores instead of nanocores)

Ability to set thresholds on node, pod, container, and cluster metrics

Kubernetes 'Labels' as indexible/searchable properties

Collection of 8 new resource types: (1) Kubernetes Sig CRD Application, (2) Deployment, (3) ReplicaSet, (4) Replication Controller, (5) DaemonSet, (6) StatefulSet, (7) Job, and (8) CronJob

7 new high level dashboards: (1) Cluster, (2) Kubernetes Sig Application CRD, (3) Deployment, (4) StatefulSet, (5) DaemonSet, (6) Job, (7) CronJob

Improved event correlation to reduce incident noise

Event propagation to view pod controller events in context of cluster, and pod events in context of service

Finer-grained event severity mapping

Separation of monitor into 2 containers, event-collector and resource-event-collector to allow finer-grained lifecycle management and collection