Prometheus is exposed on an internal endpoint that is accessible only within the cluster at. The monitoring solution is installed in a separate namespace called monitoring and uses a StorageClass called monitoring-sc. The -storageSizeGB parameter sets the size of the persistent volume that's provisioned to store metrics, and the -retentionTimeHours parameter sets the amount of time the collected metrics are retained. Run the Install-AksHciMonitoring command to deploy the monitoring solution on an existing workload cluster, as shown below: Install-AksHciMonitoring -Name mycluster -storageSizeGB 100 -retentionTimeHours 240 Option 2: Deploy the monitoring solution on an existing workload cluster The retention time for collected metrics ( retentionTimeHours) is 240 hours (or 10 days).The size of the persistent volume that's provisioned to store metrics ( storageSizeGB) is 100 GB.Monitoring is installed with the following default configuration: To enable monitoring, provide the -enableMonitoring parameter when you use New-AksHciCluster to create the workload cluster, as shown in the following example: New-AksHciCluster -name mynewcluster -enableMonitoring Option 1: Deploy the monitoring solution when creating the workload cluster This section describes the two options you can use to deploy monitoring on a workload cluster. Deploy monitoring solution using PowerShell To view the Grafana dashboards available in AKS hybrid, see Grafana dashboards available in AKS hybrid. Windows exporter to collect metrics for Windows nodes.Node exporter to collect metrics for Linux nodes.Kube state metrics to collect various metrics provided by Kubernetes.The objects and endpoints that the Prometheus solution scrapes include the following items: Since the Prometheus operator, Prometheus, and Kube state metrics components are only supported on Linux, you must provision at least one Linux node in your AKS cluster to deploy this solution. Once the Prometheus solution is deployed, the Node exporter runs on each Linux node, and the Windows exporter runs on each Windows node. The deployment is based on the publicly available Kube-Prometheus-stack helm chart, which is extended to support the Windows exporter and secures metrics scraping between Prometheus and agents. You must have your own licensed copy of Grafana with AKS hybrid.Īs part of Prometheus solution in AKS hybrid, the following components are deployed and automatically configured: You can also configure Grafana to use Prometheus as the data source. Grafana is a tool used to view, query, and visualize metrics on the Grafana dashboards. AKS hybrid makes it easy to deploy Prometheus, which is described later in this topic. Prometheus works with different types of collectors and agents to collect metrics and store them in a database where you can query the data and view reports. Prometheus is a monitoring and alerting toolkit you can use for monitoring containerized workloads. ![]() The recommended monitoring solution includes the following two tools: It's crucial that you monitor the health, performance, and resource usage of the control plane nodes and workloads on your cluster when running apps in production. To use Azure Monitor with AKS clusters, see the Azure Monitor overview. See Monitor clusters using on-premises monitoring. ![]() Supported as open-source software by Microsoft (with no support agreement or SLAs), the community, and/or external vendors.Ĭustomer-driven. ![]() Requires signing up for the Azure Monitor service. Requires connecting the AKS cluster to Azure using Azure Arc for Kubernetes.įull support and servicing from Microsoft. Two types of monitoring and logging solutions are available, as described in the following table. You'll learn how to set up and access on-premises monitoring using Prometheus and Graphana, and how to collect and view logs using Elasticsearch, Fluent Bit, and Kibana (EFK). This article describes how to monitor your Azure Kubernetes Service (AKS) deployment and collect logging data in AKS hybrid. Applies to: AKS on Azure Stack HCI, AKS on Windows Server
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |