Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Azure Database for PostgreSQL flexible server includes a high availability (HA) health status monitoring feature that uses Azure's Resource Health Check (RHC) framework. This service provides continuous insights into the health of HA-enabled instances and notifies you of events that might affect connectivity and availability. The following sections describe each health state and associated scenarios to help you troubleshoot and maintain HA stability.
Health states
The service monitors each HA state through various internal signals that represent specific conditions. The following list describes the possible HA states along with visual indicators and scenarios that might affect your Azure Database for PostgreSQL flexible server.
Ready – HA is healthy
The Ready status indicates that your HA-enabled server is operating normally with no detected issues that affect failover readiness. All necessary configurations are intact, and no significant error conditions exist.
Degraded – Network Security Group (NSG) or virtual appliance blocking connections
The Degraded status might appear when NSG rules or a virtual appliance block essential connections required for high availability. This configuration issue prevents full HA functionality. Correct the issue by adjusting the NSG settings.
Degraded – Read-only state
If your PostgreSQL flexible server enters a read-only state, the Degraded status reflects this restriction. This state typically requires provisioning additional resources or addressing the conditions that led to the read-only setting to restore full functionality.
Degraded – High availability in degraded state
This status appears when the high availability (HA) service experiences degraded performance, possibly due to transient issues or system-level conditions. Implementing retry logic can help mitigate the effects of these temporary connectivity disruptions. The degraded status doesn't mean the server is unavailable. Instead, it indicates that the overall HA setup and health checks aren't fully complete. Despite this status, the server might still be operational and accessible.
To accurately monitor your database's availability during such periods, use the is_db_alive metric as part of Database Availability Metrics. This metric provides a reliable indicator of the database's availability, helping you distinguish between a temporary incomplete HA setup and actual downtime.
Degraded – Planned failover initiated
During a planned failover event initiated for your server, the Degraded status appears, signifying that HA failover processes are active. This process is generally brief and controlled, and service should resume shortly.
Degraded – Unplanned failover initiated
For an unplanned failover, this status indicates an active failover event triggered by unexpected circumstances. This scenario might involve brief connectivity interruptions until the server completes failover procedures.
Degraded – Upgrade failover initiated
During system upgrades, your HA server might undergo an upgrade failover to apply necessary updates. While in this state, the server might restrict new connections temporarily. Implement retry logic to handle transient issues effectively.
Configuring Resource Health Alerts
Set up Resource Health alerts to get real-time notifications when any changes occur in the health status of your HA-enabled PostgreSQL instance. You can configure these alerts through the Azure portal or by using an ARM template. This way, you stay informed of HA status updates without actively monitoring the portal.
Steps to Configure Resource Health Alerts via Portal
- Go to the Azure portal and select your PostgreSQL flexible server.
- In the left-hand menu, under the Monitoring section, select Alerts.
- Select New alert rule and configure the alert logic based on Resource Health signals.
- Set up the action group to specify how you want to be notified (email, SMS, and so on).
- Review and create the alert rule.
Steps to Create Resource Health Alerts using ARM Template
- Download the ARM template from the Resource Health Alerts ARM Template Guide.
- Customize the template with your specific server details and alert preferences.
- Deploy the ARM template by using Azure CLI or Azure PowerShell.
- Verify the deployment and ensure the alerts are active.
For more details on setting up alerts, see the following guides:
By using HA Health Status Monitoring, you gain essential insights into your PostgreSQL server's HA performance, enabling a proactive approach to managing uptime and availability.