
上QQ阅读APP看书,第一时间看更新
Monitoring Azure Stack
As Azure Stack is an integrated system, the monitoring solution for Azure Stack itself (and not the tenant workloads) has two pillars:
- Hardware monitoring
- Azure Stack monitoring
Hardware monitoring: As the Azure Stack hardware is vendor based, the monitoring of the components works with the hardware vendor-based monitoring solutions that already exist and have been proofed in the based without Azure Stack itself:

Azure Stack monitoring: The Azure Stack software monitoring is based on the corresponding health APIs. If you are running System Center Operations Manager (SCOM) on-premise, a SCOM solution is available to monitor multiple Azure Stacks with one SCOM:

If your primary monitoring solution is Nagios based, there is an integration available, too.