In this blog, I previewed some new features in Foglight for Virtualization Standard Edition 7.0. Now that FVS 7.0 is generally available, let’s look at two other key new features, Power Minimization and Host Refresh.
A new Power Minimization feature has been added to the new Planning tab in Capacity Manager:
In the chart above, FVS analyzes workloads across the selected existing resource object (in the far left pane), and consolidates the VMs onto the fewest hosts possible while ensuring the system has enough capacity to meet SLAs while maintaining fault tolerance and high availability policies. This chart provides a historical view of the minimum number of hosts needed for the workloads during the past 24 hour period, represented by the blue line.
We can change the historical period in the gray Selectors box. In this example we’re analyzing our entire infrastructure including VMware, Microsoft, and Red Hat environments. In this example we’re showing a single line representing the minimum number of hosts needed across multiple clusters. We can show more granular information by displaying a line for each clusters by unchecking the Combine Clusters box at the bottom of the chart. This will show us the minimum number of hosts needed in each cluster over the selected time period.
In this example, in our entire environment if we were to consolidate VMs onto the fewest possible hosts and power down unneeded hosts over time (while powering up hosts as needed when business demand requires it), we would reduce power usage by 16%. Since part of the environment includes VMware clusters, this process can be automated through VMware’s Distributed Power Management feature.
We can also model workloads against potential new host server hardware by clicking the New Server radio button in the gray Selector box. This can help us analyze potential before and after scenarios for refresh projects.
Also new in the Planning tab of Capacity Manager is a Host Refresh feature:
In Host Refresh, we can enter new host server hardware models, including their resource configuration, cost, and space and power requirements by clicking the Add button in the main window. Once we’ve input our servers, we can analyze existing and planned workloads against the new hardware and determine which hardware will best meet our needs based on weighting factors we can apply to server cost, power, and space requirements. In this example above, we’ve selected our entire infrastructure spanning VMware, Microsoft, and Red Hat clusters and are modeling all existing workloads against three potential new Dell PowerEdge servers.
In the analysis, FVS models existing workloads while accounting for cluster fault tolerance and failover policies against each server and stack ranks them based on the weighting factors. We’re applying a 100% weight to server cost and ignoring power and space requirements in this example. The Ranking column in the table displays the rank, while the Requirement column shows the minimum number of servers we need in this environment to safely execute all workloads to meet SLAs and cluster policies. In this example it turns out the lowest cost server our needs best, but in many environments that won’t be the case.
We can generate reports of analyses in both of these tabs by clicking the Report button just under the tabs.
Visit https://www.quest.com/products/foglight-for-virtualization-enterprise-edition/ for more information.