This week Blue Medora is releasing the 3rd major release of our Oracle Enterprise Manager plugin for VMware. Over the past year our customers have been telling us that, while they love the deep health, availability, and performance metrics and incidents the VMware plugin adds to their EM12c environments, they also desired the ability to “manage” the VMware virtual machines that their Oracle workloads are running on in a similar fashion that Oracle itself provides for Oracle VM within EM. This latest release includes a slew of major functional enhancements that deliver on those “manage VMware from Oracle EM” requirements. In the next few weeks we’ll be drilling into the new features in more detail, however, we wanted to outline the functionality we are most excited about:
One of the most common interactions between Oracle teams and infrastructure teams are hardware requirement change requests. It may be as simple as requesting an additional GB of RAM, however that request often becomes bogged down by process and take days, weeks or even sometimes months to accomplish. The latest release of the Plugin for VMware includes virtual machine scaling capabilities. Leveraging the Enterprise Manager jobs framework, we added new functionality that allows you adjust a virtual machine’s settings directly in the EM12c.
One of the more advanced and popular use cases of this functionality include tying the scaling operating to metric collections as a corrective action. For example, you can define a policy such that if a virtual machine has 5 consecutive collections of high CPU utilization, you can add CPU.
Exactly as it sounds — this release adds the capability of provisioning new VMware virtual machines from VMware templates or existing virtual machines. With this capability it is no longer an absolute requirement to have to involve the VMware IT infrastructure team in the provisioning process when VMware VMs need to be brought online to support additional Oracle workloads. It’s now all possible directly from within EM12c.
When we ask Oracle DBAs what their #1 challenge is in managing their Oracle workloads virtualized on VMware their answer is consistently related to storage issues. When virtualizing, the storage layer is now abstracted a layer further away that it had been ith traditional deployments on bare metal. With the latest release of the VMware plugin, Blue Medora has included a slew additional storage metrics as well as a new dedicated VMware vSphere Datastore EM target type to provide the insights DBAs and other Oracle admins need to identify and isolated storage related performance bottlenecks.
From a disk perspective, the Plugin for VMware includes metrics such as read, writes, throughput, commands aborted, and those ever-important IOPs.
The EM12c-based “discovery” of all of the VMware infrastructure components (ESX Hosts, Datastores, VMs, Data Centers, Clusters, etc) has been dramatically improved and will reduce both the number of steps and the amount of time it takes to install and configure the VMware plugin. We also expanded a customers ability to set thresholds for alerting at each of the VMware target types — providing a more fine grained alerting capability.
In previous releases all the thresholds had to be defined on the VMware vSphere high level EM target type. Finally, we’ve changed the UI paradigm in a significant way in that we now allow EM administrators to ‘scope’ the views in the plugin to only the VMware virtual machines they’ve actually promoted in the EM discovery process. This new approach dramatically reduces the amount of “noise” Oracle admins have to sift through to identify, troubleshoot, and generally monitor Oracle workloads running on VMware.