IT Operations Management (ITOM)
cancel

4 keys to monitoring cloud and making it easy with HP vPV

4 keys to monitoring cloud and making it easy with HP vPV

Ops_Guest

Guest post by Manish Jain, Software Developer, HP Virtualization Performance Viewer

 

Choosing a monitoring solution for your cloud is no easy task. It involves a lot of complexity. Here are some points to ponder when it comes to choosing a monitoring solution:

  1. A typical cloud environment can make use of multiple hypervisor technologies to provision cloud services, various storage devices for provisioning storage and of course an extensive network infrastructure.
  2. There are multiple vendors for each of these cloud components (compute, storage, networking etc.). So the monitoring solution should be vendor neutral.
  3. A cloud environment could have thousands of virtual machines, hypervisor hosts and many other infrastructure elements—all of which will generate enormous amount of performance data. So the monitoring solution needs to be built with an architecture that can consume large amounts of data.
  4. Once you have this amount of data gathered over a time period, the solution should have an analytical component which can provide actionable recommendations to proactively avoid performance bottlenecks, optimize the environment for efficiency, and plan for future infrastructure capacity needs.

A cloud administrator requires a unified tool for performance monitoring, capacity optimization, analysis and planning. Plus the tool should be easy to install and maintain. HP Virtualization Performance Viewer (vPV) is this type of tool.

 

A cloud administrator can start monitoring the environment with vPV in minutes. The administrator needs to register the OpenStack cloud controller and hypervisor data source (KVM, VMware vCenter Server) with vPV. Once registered, vPV will start collecting performance data from these data-source targets.

 

Moreover, vPV performs the management of your cloud resources remotely. It does not require you to install any additional software component on your cloud resource. It uses REST APIs of OpenStack components (Nova, Cinder etc.) to get the topology information like controller, tenants, hypervisors, instances, etc. vPV uses its data collectors of different virtualization technologies (VMware, KVM etc.) to source the performance data from virtualization platform. The topology data collected from OpenStack and performance data collected from virtualization platform is then reconciled by vPV. This allows use of in-depth data collection of vPV for different hypervisors like KVM, VMWare, Xen, and MS Hyper-V, instead of relying on any tool that provides only few common metrics across all hypervisors.

 

Figure 1 – Monitoring OpenStack with vPV

 

 

Now after collecting topology, configuration and utilization data, vPV analyzes the capacity, usage, and allocation trends for various resources in a cloud environment. It displays various analysis and reports that makes it easy for the administrator to detect performance anomalies and rectify the issue on a specific resource/entity.

HP vPV provides capacity optimization, what-if modeling, forecasting and performance triaging capabilities. These capabilities helps admin to make decisions like:

  • Capacity Modeler - As an Administrator, you may want to analyze how the allocation of resources impacts your environment. It is important to get accurate results from the analysis. Based on the analysis, you can plan and allocate the infrastructure capacity effectively.

 

  • Placement - This capability provides recommendations as which is the optimal cloud resource pool (cluster) to spawn VM or VMs based on current capacity utilization and demand trends.

 

  • Optimization – Recommendations to right-size the VMs to release unused resources and improve performance.

 

  • Forecasting - Based on historical usage,vPV can forecast the resource utilization of entities in the environment. You can use these projections to plan for future resource requirements and allocations.

 

  • Performance Monitoring – Quickly identify hotspots and perform triage historical analysis across several infrastructure resources with in-depth and extensive performance counters.

 

For more information on how Virtualized Performance Viewer can help with your cloud and virtualization monitoring needs visit the homepage here. vPV will also be showcased at HP Discover Las Vegas at the HP Software/BSM arena. You can register for HP Discover here

 

You can also learn more about vPV in this blog:

  • infrastructure management
About the Author

Ops_Guest