From 6c33804ab0de9da48e798abcd3b63bf81bb684c0 Mon Sep 17 00:00:00 2001 From: Catherine Luse Date: Sun, 11 Apr 2021 11:01:06 -0700 Subject: [PATCH] Add docs for Windows cluster support for Monitoring v2 --- .../v2.5/en/monitoring-alerting/_index.md | 11 ++++ .../windows-clusters/_index.md | 57 +++++++++++++++++++ 2 files changed, 68 insertions(+) create mode 100644 content/rancher/v2.5/en/monitoring-alerting/windows-clusters/_index.md diff --git a/content/rancher/v2.5/en/monitoring-alerting/_index.md b/content/rancher/v2.5/en/monitoring-alerting/_index.md index 611c9242132..92b95771eff 100644 --- a/content/rancher/v2.5/en/monitoring-alerting/_index.md +++ b/content/rancher/v2.5/en/monitoring-alerting/_index.md @@ -31,6 +31,7 @@ For more information about upgrading the Monitoring app in Rancher 2.5, please r - [About Prometheus](#about-prometheus) - [Enable Monitoring](#enable-monitoring) - [Default Alerts, Targets, and Grafana Dashboards](#default-alerts-targets-and-grafana-dashboards) +- [Windows Cluster Support](#windows-cluster-support) - [Using Monitoring](#using-monitoring) - [Grafana UI](#grafana-ui) - [Prometheus UI](#prometheus-ui) @@ -84,6 +85,16 @@ To see the default dashboards, go to the [Grafana UI.](#grafana-ui) In the left To configure Prometheus resources from the Rancher UI, click **Apps & Marketplace > Monitoring** in the upper left corner. +# Windows Cluster Support + +_Available as of v2.5.8_ + +Monitoring v2 can now be deployed on Windows nodes. + +To be able to fully deploy Monitoring V2 for Windows, all of your Windows hosts must have a minimum wins version of v0.1.0. + +For more details on how to upgrade wins on existing Windows hosts, refer to the section on [Windows cluster support for Monitoring v2.](./windows-clusters) + # Using Monitoring Installing `rancher-monitoring` makes the following dashboards available from the Rancher UI. diff --git a/content/rancher/v2.5/en/monitoring-alerting/windows-clusters/_index.md b/content/rancher/v2.5/en/monitoring-alerting/windows-clusters/_index.md new file mode 100644 index 00000000000..10d06acfe34 --- /dev/null +++ b/content/rancher/v2.5/en/monitoring-alerting/windows-clusters/_index.md @@ -0,0 +1,57 @@ +--- +title: Windows Cluster Support for Monitoring v2 +shortTitle: Windows Clusters +weight: 5 +--- + +_Available as of v2.5.8_ + +Starting at Monitoring V2 9.4.204 (used by default in Rancher 2.5.8), Monitoring v2 can now be deployed on a Windows cluster and will scrape metrics from Windows nodes using [windows_exporter](https://github.com/prometheus-community/windows_exporter) (a community project previously named `wmi_exporter` when deployed in Monitoring V1). + +- [Comparison to Monitoring v1](#comparison-to-monitoring-v1) +- [Cluster Requirements](#cluster-requirements) + - [Upgrading Existing Clusters to wins v0.1.0](#upgrading-existing-clusters-to-wins-v0-1-0) + +# Comparison to Monitoring v1 + +Unlike Monitoring v1 for Windows, metrics collected by `windows_exporter` will be labeled as `windows_` instead of `wmi_` in accordance to a naming change from upstream from `wmi_exporter` to `windows_exporter`. + +In addition, Monitoring V2 for Windows will leverage a new feature in `wins` v0.1.0 to support scraping `windows_exporter` metrics from the Pod itself rather than from the Node. This means that users will no longer have to keep port 9796 open on each Windows host to allow Prometheus to scrape metrics. + +# Cluster Requirements + +Monitoring V2 for Windows can only scrape metrics from Windows hosts that have a minimum `wins` version of v0.1.0. To be able to fully deploy Monitoring V2 for Windows, all of your hosts must meet this requirement. + +If you provision a fresh cluster in Rancher 2.5.8, your cluster should already meet this requirement. + +### Upgrading Existing Clusters to wins v0.1.0 + +If the cluster was provisioned before Rancher 2.5.8 (even if the current Rancher version is 2.5.8), you will not be able to successfully deploy Monitoring V2 for Windows until you upgrade the wins version on each host to at least v0.1.0. + +To facilitate this upgrade, Rancher 2.5.8 has released a brand new Helm chart called `rancher-wins-upgrader`. + +> **Prerequisite:** Make sure Monitoring v1 for Windows is uninstalled. + +1. Deploy `rancher-wins-upgrader` with the following override: + ```yaml + # Masquerading bootstraps the wins-upgrader installation via + # a previously whitelisted process path since the normal install path, + # c:\etc\rancher\wins\wins-upgrade.exe is not normally whitelisted. + # In this case, we are using the previously whitelisted process + # path used by Monitoring V1. + masquerade: + enabled: true + as: c:\\etc\wmi-exporter\wmi-exporter.exe + ``` + > **Note for Non-Default Windows Prefix Path:** If you set up the RKE cluster with a `cluster.yml` that has a non-default `win_prefix_path`, you will need to update the `masquerade.as` field with your prefix path in place of `c:\\`. + > + > For example, if you have `win_prefix_path: 'c:\host\opt\'`, then you will need to set `as: c:\host\opt\etc\wmi-exporter\wmi-exporter.exe`. +2. Once all your hosts have been successfully upgraded, please ensure that you deploy the Helm chart once again with default values to avoid conflicts with the following settings: + ```yaml + masquerade: + enabled: false + ``` + +**Result:** The hosts are ready for Monitoring v2 to be installed. You may choose to uninstall the `rancher-wins-upgrader` chart or keep it in your cluster to facilitate future upgrades. + +For more information on how it can be used, please see the [README.md](https://github.com/rancher/wins/blob/master/charts/rancher-wins-upgrader/README.md) of the chart.