Merge pull request #300211 from rdeltcheva/az-events-az-upd

Expand azure-events-az explanation

Author: prmerger-automator[bot]

articles/sap/workloads/get-started.md

@@ -55,6 +55,7 @@ In the SAP workload documentation space, you can find the following areas:
 
 ## Change Log
 
+- May 21, 2025: Change in [Set up Pacemaker on RHEL in Azure](./high-availability-guide-rhel-pacemaker.md) and [Set up Pacemaker on SLES in Azure](./high-availability-guide-suse-pacemaker.md) to expand the explanation how azure-events-az works.
 - April 29, 2025: Updated the ENSA2 order constraint in the RHEL pacemaker setup in [High availability for SAP NetWeaver on VMs on RHEL with NFS on Azure Files](./high-availability-guide-rhel-nfs-azure-files.md), [Azure Virtual Machines HA for SAP NetWeaver on RHEL with Azure NetApp Files for SAP applications](./high-availability-guide-rhel-netapp-files.md), [Azure Virtual Machines high availability for SAP NetWeaver on Red Hat Enterprise Linux](./high-availability-guide-rhel.md), and [High availability for SAP NetWeaver on Azure VMs on Red Hat Enterprise Linux for SAP applications multi-SID](./high-availability-guide-rhel-multi-sid.md).
 - April 9, 2025: Add timeout and fail action to IP resource configuration for SAP HANA in [High availability for SAP HANA on Azure VMs on SUSE Linux Enterprise Server](./sap-hana-high-availability.md) and [High availability for SAP HANA scale-out system with HSR on SUSE Linux Enterprise Server](./sap-hana-high-availability-scale-out-hsr-suse.md)
 - March 31, 2025: Updated Azure NetApp Files Application Volume Group guidelines for SAP HANA in [FAQs for Virtual Machine Scale Set for SAP workload](./virtual-machine-scale-set-sap-faq.md)

articles/sap/workloads/high-availability-guide-rhel-pacemaker.md

@@ -907,11 +907,13 @@ The monitoring and fencing operations are deserialized. As a result, if there's
 
 ## Configure Pacemaker for Azure scheduled events
 
-Azure offers [scheduled events](/azure/virtual-machines/linux/scheduled-events). Scheduled events are sent via the metadata service and allow time for the application to prepare for such events.
+Azure offers [scheduled events](/azure/virtual-machines/linux/scheduled-events). Scheduled events are provided via the metadata service and allow time for the application to prepare for such events. 
 
-The Pacemaker resource agent `azure-events-az` monitors for scheduled Azure events. If events are detected and the resource agent determines that another cluster node is available, it sets a cluster health attribute.
+Resource agent [azure-events-az](https://github.com/ClusterLabs/resource-agents/pull/1161) monitors for scheduled Azure events. If events are detected and the resource agent determines that another cluster node is available, it sets a node-level health attribute `#health-azure` to `-1000000`. 
 
-When the cluster health attribute is set for a node, the location constraint triggers and all resources with names that don't start with `health-` are migrated away from the node with the scheduled event. After the affected cluster node is free of running cluster resources, the scheduled event is acknowledged and can execute its action, such as a restart.
+When this special cluster health attribute is set for a node, the node is considered unhealthy by the cluster and all resources are migrated away from the affected node. The location constraint ensures resources with name starting with ‘health-‘ are excluded, as the agent needs to run in this unhealthy state. Once the affected cluster node is free of running cluster resources, scheduled event can execute its action, such as restart, without risk to running resources. 
+
+The `#heath-azure` attribute is set back to `0` on pacemaker startup once all events have been processed, marking the node as healthy again.
 
 1. **[A]** Make sure that the package for the `azure-events-az` agent is already installed and up to date.
 

articles/sap/workloads/high-availability-guide-suse-pacemaker.md

@@ -8,7 +8,7 @@ ms.service: sap-on-azure
 ms.subservice: sap-vm-workloads
 ms.topic: article
 ms.custom: devx-track-azurepowershell, linux-related-content
-ms.date: 08/26/2024
+ms.date: 05/21/2025
 ms.author: radeltch
 ---
 
@@ -1000,7 +1000,13 @@ sudo crm configure property stonith-timeout=900
 
 ## Configure Pacemaker for Azure scheduled events
 
-Azure offers [scheduled events](/azure/virtual-machines/linux/scheduled-events). Scheduled events are provided via the metadata service and allow time for the application to prepare for such events. Resource agent [azure-events-az](https://github.com/ClusterLabs/resource-agents/pull/1161) monitors for scheduled Azure events. If events are detected and the resource agent determines that another cluster node is available, it sets a cluster health attribute. When the cluster health attribute is set for a node, the location constraint triggers and all resources, whose name doesn't start with "health-" are migrated away from the node with scheduled event. Once the affected cluster node is free of running cluster resources, scheduled event is acknowledged and can execute its action, such as restart.
+Azure offers [scheduled events](/azure/virtual-machines/linux/scheduled-events). Scheduled events are provided via the metadata service and allow time for the application to prepare for such events. 
+
+Resource agent [azure-events-az](https://github.com/ClusterLabs/resource-agents/pull/1161) monitors for scheduled Azure events. If events are detected and the resource agent determines that another cluster node is available, it sets a node-level health attribute `#health-azure` to `-1000000`. 
+
+When this special cluster health attribute is set for a node, the node is considered unhealthy by the cluster and all resources are migrated away from the affected node. The location constraint ensures resources with name starting with ‘health-‘ are excluded, as the agent needs to run in this unhealthy state. Once the affected cluster node is free of running cluster resources, scheduled event can execute its action, such as restart, without risk to running resources. 
+
+The `#heath-azure` attribute is set back to `0` on pacemaker startup once all events have been processed, marking the node as healthy again.
 
 > [!IMPORTANT]
 > Previously, this document described the use of resource agent [azure-events](https://github.com/ClusterLabs/resource-agents/blob/main/heartbeat/azure-events.in). New resource agent [azure-events-az](https://github.com/ClusterLabs/resource-agents/blob/main/heartbeat/azure-events-az.in) fully supports Azure environments deployed in different availability zones.