Virtualization is supposed to do so much more than just consolidate servers. It was supposed to allow IT to become more service oriented, essentially becoming the organization’s cloud, provisioning IT resources on the fly. Moving beyond simple server consolidation to a cloud requires monitoring, management and orchestration, and the reality is many organizations have only taken very small steps in that direction.
Time to End Over-Provisioning
Most IT designs are done by over-provisioning the various resources the environment needs, making sure that under a worst case peak in demand the environment has more than enough compute, network and storage resources to continue to deliver to the organization’s expectations.
The problem with over-provisioning is that most of the time the required resources go unused. A better strategy, and one that truly leverages virtual machine mobility, is to design the data center for normal levels of performance demand.
The danger in designing to the norm is how to handle a situation where there is a spike in performance on a particular workload. Virtualization enables workload mobility, so moving the workload to a physical host with more available compute and networking than its current physical host certainly is possible. The problem is how does IT, already stretched too thin, anticipate these peaks prior to them impacting the user experience?
Anticipation Needs To Be Automated
For almost every data center, there are too many workloads with too many interdependent virtual machines for a single IT administrator, or even a team of administrators, to proactively monitor and anticipate potential problems. The process needs to be driven automatically through software.
VMware’s solution to the problem is vRealize Operations. It helps IT professionals plan, manage and scale their environment fully realizing the full potential of the software defined data center. The solution enables the management of IT operations from applications to infrastructure. But performance as a sum of more than just the efficiency of the application and effectiveness of the compute. The network and storage infrastructure play a critical role in overall performance.
Peering into Storage
VMware has implemented an extendable framework into VRealize Operations, enabling networking and storage vendors to integrate into the interface. The problem is these integrations are unique to each vendor, and while they plug into the same interface as vRealize, they don’t always cross communicate between storage systems. The information is still siloed.
Introducing Blue Medora True Visibility
Blue Medora enables the consolidation of multiple IT operations tools into a single solution. Blue Medora has an extensive list of storage hardware that it can monitor and report on. It can then cross-map the information from the devices it supports directly into vRealize Operations. True Visibility integrates directly with the vRealize Operations dashboards for an instant assessment of workload health and its performance.
Blue Medora extensions to vRealize enable the solution to become a end-to-end monitoring and management tool for VMware environments. It provides drill down to identify noisy neighbors, bandwidth consumers and poor queries the impact VM performance.
Just The Right Amount of Information
Most data centers have tried various monitoring and management tools. The problem is most of these tools provide too much information, and the “right” information is lost in the noise. With vRealize Operations and True Visibility working together to ensure it flags only the most important notifications.
The other problem with most solutions is that they do alert of a potential problem, but provide no advice on what to do to fix the problem. With True Visibility the IT administrator is provided with a series of recommendations that help them address root causes and make sure the problem does not occurring the future.
Finally, being end-to-end also means the solution needs to be dependency aware. Most applications are made up of a group of VMs, multiple network paths and multiple storage volumes. True Visibility is dependency aware. It understands the relationship between multiple VMs, the networks they reside on and the storage they require.
A true private cloud is one that automatically adapts as workload performance expectations increase or decrease. In IT compute, network and storage are seen as a pools of resources that can be allocated on-the-fly to workloads when needed and then release by those workloads when the peak passes. But to attain this state requires tools like vRealize Operations and the Blue Medora management packs so changes can be predicted and addressed prior to impacting performance.