Service mesh component resource custom configuration¶
This page describes how to customize mesh component resources via Container Management. The control plane components of the service mesh are as follows:
Component Name | Location | Description | Default Resource Settings |
---|---|---|---|
mspider-ui | Global Management Cluster | Service Mesh Interface | requests: CPU: Not set; Memory: Not set limits: CPU: Not set; Memory: Not set |
mspider-ckube | Global management cluster | Acceleration component of Kubernetes API Server, used to call global cluster-related resources | requests: CPU: not set; memory: not set limits: CPU: not set; memory: not set settings |
mspider-ckube-remote | Global Management Cluster | Used to call Kubernetes of remote clusters, aggregate multicluster resources, and accelerate | requests: CPU: not set; memory: not set limits: CPU: not set; memory : not set |
mspider-gsc-controller | Global management cluster | Service mesh management component, used for mesh creation, mesh configuration and other mesh control plane lifecycle management, and Mspider control plane capabilities such as permission management | requests: CPU: Not set ;memory: not set limits: CPU: not set; memory: not set |
mspider-api-service | Global management cluster | Provide interface for Mspider background API interaction and other control behaviors | requests: CPU: not set; memory: not set limits: CPU: not set; memory: not set |
Hosted mesh | |||
istiod-{meshID}-hosted | control plane cluster | policy management for hosted mesh | requests: CPU: 100m; memory: 100m limits: CPU: not set; memory: not set |
mspider-mcpc-ckube-remote | Control plane cluster | Invoke remote mesh work clusters to accelerate and aggregate multicluster resources | requests: CPU: 100m; memory: 50m limits: CPU: 500m; memory: 500m |
mspider-mcpc-mcpc-controller | Control plane cluster | Aggregate mesh multicluster related data plane information | requests: CPU: 100m; memory: 0 limits: CPU: 300m; memory: 1.56G |
{meshID}-hosted-apiserver | Control Plane Cluster | Hosted Control Plane Virtual Cluster API Server | requests: CPU: not set; memory: not set limits: CPU: not set; memory: not set |
istiod | working cluster | Mainly used for sidecar lifecycle management of the cluster | requests: CPU: 100; memory: 100 limits: CPU: not set; memory: not set |
Proprietary mesh | |||
istiod | used for policy creation, delivery, and sidecar lifecycle management | requests: CPU: 100; memory: 100 limits: CPU: not set; memory: not set | |
mspider-mcpc-ckube-remote | working cluster | call remote mesh working cluster | requests: CPU: 100m; memory: 50m limits: CPU: 500m; memory: 500m |
mspider-mcpc-mcpc-controller | Working cluster | Collect cluster data surface information | requests: CPU: 100m; memory: 0 limits: CPU: 300m; memory: 1.56G |
External mesh | |||
mspider-mcpc-ckube-remote | working cluster | call remote mesh working cluster | requests: CPU: 100m; memory: 50m limits: CPU: 500m; memory: 500m |
mspider-mcpc-mcpc-controller | Working cluster | Collect cluster data surface information | requests: CPU: 100m; memory: 0 limits: CPU: 300m; memory: 1.56G |
The preset resource settings of each control plane component of the service mesh are shown in the above table. Users can find the corresponding workload in the [Container Management] module and customize CPU and memory resources for the workload.
Prerequisites¶
The cluster has been managed by the service mesh, and the mesh components have been installed normally; The login account has the admin or editor authority of the namespace istio-system in the global management cluster and the working cluster;
set operation¶
Take istiod on the working cluster under the hosted mesh as an example, the specific operations are as follows:
-
View the hosted mesh nicole-dsm-mesh access cluster under the service mesh is nicole-dsm-c2, as shown in the figure below.
-
Click the cluster name, jump to the cluster page in the Container Management module, click to enter the Workload -> Stateless Load page to find istiod;
-
Click the workload name to enter Container Configuration -> Basic Information tab page;
-
Click the Edit button to modify the CPU and memory quotas, click Next , OK .
-
View the Pod resource information under the workload, and it can be seen that it has changed.