Enable Pluggable Components — Overview

KubeSphere has decoupled some core feature components since v2.1.0. These components are designed to be pluggable which means you can enable them either before or after installation. By default, KubeSphere will be deployed with a minimal installation if you do not enable them.

Different pluggable components are deployed in different namespaces. You can enable any of them based on your needs. It is highly recommended that you install these pluggable components to discover the full-stack features and capabilities provided by KubeSphere.

For more information about how to enable each component, see respective tutorials in this chapter.

Resource Requirements

Before you enable pluggable components, make sure you have enough resources in your environment based on the tables below. Otherwise, components may crash due to a lack of resources.

Note

The following request and limit of CPU and memory resources are required by a single replica.

KubeSphere App Store

Namespaceopenpitrix-system
CPU Request0.3 core
CPU LimitNone
Memory Request300 MiB
Memory LimitNone
InstallationOptional
NotesProvide an App Store with application lifecycle management. The installation is recommended.

KubeSphere DevOps System

Namespacekubesphere-devops-systemkubesphere-devops-system
PatternAll-in-One installationMulti-node installation
CPU Request34 m0.47 core
CPU LimitNoneNone
Memory Request2.69 G8.6 G
Memory LimitNoneNone
InstallationOptionalOptional
NotesProvide one-stop DevOps solutions with Jenkins pipelines and B2I & S2I.The memory of one of the nodes must be larger than 8 G.

KubeSphere Monitoring System

Namespacekubesphere-monitoring-systemkubesphere-monitoring-systemkubesphere-monitoring-system
Sub-component2 x Prometheus3 x AlertmanagerNotification Manager
CPU Request100 m10 m100 m
CPU Limit4 coresNone500 m
Memory Request400 MiB30 MiB20 MiB
Memory Limit8 GiBNone1 GiB
InstallationRequiredRequiredRequired
NotesThe memory consumption of Prometheus depends on the cluster size. 8 GiB is sufficient for a cluster with 200 nodes/16,000 Pods.--

Note

The KubeSphere monitoring system is not a pluggable component. It is installed by default. The resource request and limit of it are also listed on this page for your reference as it is closely related to other components such as logging.

KubeSphere Logging System

Namespacekubesphere-logging-systemkubesphere-logging-systemkubesphere-logging-systemkubesphere-logging-system
Sub-component3 x Elasticsearchfluent bitkube-eventskube-auditing
CPU Request50 m20 m90 m20 m
CPU Limit1 core200 m900 m200 m
Memory Request2 G50 MiB120 MiB50 MiB
Memory LimitNone100 MiB1200 MiB100 MiB
InstallationOptionalRequiredOptionalOptional
NotesAn optional component for log data storage. The internal Elasticsearch is not recommended for the production environment.The log collection agent. It is a required component after you enable logging.Collecting, filtering, exporting and alerting of Kubernetes events.Collecting, filtering and alerting of Kubernetes and KubeSphere auditing logs.

KubeSphere Alerting and Notification

Namespacekubesphere-alerting-system
CPU Request0.08 core
CPU LimitNone
Memory Request80 M
Memory LimitNone
InstallationOptional
NotesAlerting and Notification need to be enabled at the same time.

KubeSphere Service Mesh

Namespaceistio-system
CPU Request1 core
CPU LimitNone
Memory Request3.5 G
Memory LimitNone
InstallationOptional
NotesSupport grayscale release strategies, traffic topology, traffic management and distributed tracing.