Technical Blog Post
IBM PureApplication System’s Monitoring Capabilities in a Nutshell
Setting up and managing an infrastructure to monitor your cloud-based IT is often a labor-intensive effort that requires time and deep technical skills. IBM PureApplication System enables you to dramatically simplify this task via its built-in monitoring support.
By taking advantage of PureApplication’s System Monitoring Shared Service, you can easily deploy a dynamically scalable client-server-agent based infrastructure on PureApplication System to monitor your workload. Depending on the operating system and the installed middleware, selected monitoring agents are injected into the virtual machines (VMs). The installation, configuration and management of this environment is fully automated and does not require any manual intervention.
Once the service is running, the Monitoring Portal can be started on the end user’s workstations to view the consolidated monitoring data. The portal also allows you to define conditions upon which alerts will be raised, e.g. when the file system allocation of a virtual machine exceeds a certain threshold.
Database Performance Monitoring is another fully-automated service to monitor DB2 database performance metrics. It allows you to isolate and analyze typical database performance problems and supports alert threshold settings to help you monitor events and conditions. Its web-based user interface is accessible from the PureApplication console. Both, System Monitoring and Database Performance Monitoring are available not only on PureApplication System, but also on IBM PureApplication Software and IBM PureApplication Service.
While the shared services’ focus is on workload monitoring, monitoring information for PureApplication’s hardware and management software can be found in the system’s console. The Infrastructure Map is a good point to start with. It provides basic information for all hardware components and includes links to additional panels with more detailed monitoring data.
A PureApplication internal events receiver collects system events from the various system-internal hardware, management and virtualization components. Depending on the configuration of the Monitoring Shared Services, this can also include alerts originating from workload. They all are reported in the console’s “Events” panel.
For critical events that require IBM support, the Service and Support Manager (Call Home Feature) allows PureApplication to automatically create a ticket (Problem Management Record) with IBM and to collect and upload the logs needed by IBM for troubleshooting.
You might have an existing monitoring infrastructure already and would like to use it to monitor PureApplication System and its workload. PureApplication supports such an integration by various means. For clients using IBM Tivoli Monitoring (ITM), the System Monitoring Shared Service can be configured to connect to an existing ITM Hub-TEMS. Workload alerts can be forwarded via the Monitoring Service's Event Integration Facility (EIF) to EIF event receivers. In addition, all kind of events collected by the PureApplication internal events receiver can be forwarded to an external destination as Simple Network Management Protocol (SNMP) traps in real-time. The picture below shows a typical setup, in which IBM Tivoli Netcool/OMNIbus serves as the SNMP trap receiver and thereby provides system events monitoring.
While this overview focuses on a single PureApplication System installation, the same concepts apply in a multi-system environment. Within a Deployment Subdomain, a single instance of the Monitoring Shared Service can be used to monitor workload that is deployed across systems.
Further Reading / Documentation:
- PureApplication System W2500 2.1 (Intel) Monitoring
- PureApplication System W2700 2.1 (Power) Monitoring
- PureApplication Service 2.1 Monitoring
- PureApplication Software 2.1 Monitoring