In previous posts in this series, I introduced the SMART framework for hyperconvergence and discussed software-only solutions on servers of choice being the best foundation for hyperconvergence. Today we will continue our discussion on hyperconvergence focusing on Management Simplicity – one of the five pillars of SMART Hyperconvergence.
Management challenges are a prominent driver for hyperconvergence
One of the factors that led to the emergence of hyperconverged solutions was the need to address operational challenges arising from server virtualization. In a virtualized environment, compute resources are usually connected to network storage. Often, one team manages the servers and virtual machines while another manages storage. The two teams must coordinate storage capacity allocation to accommodate for compute resources required for running applications.
That coordination effort is often challenging, especially in large environments, since many considerations around capacity planning, replication, data protection and performance must be taken into account. And since there is a fundamental mismatch between storage constructs (volumes & LUNs) and those of the virtualized environments (virtual machines) – that complexity has an impact on every part of the application lifecycle – provisioning, cloning, backup, scaling, etc.
In a hyperconverged environment, however, things are much simpler. There are no volumes to provision or LUNs to configure, and no need to map them to the VMs we’d like to provision. Provisioning capacity translates to simply creating a datastore and placing VMs on that datastore. Therefore, eliminating network storage effectively removes much of the inherit storage complexity associated with these traditional arrays.
With a single resource pool that is aware of the constructs running on top of it – in this case virtual machines – things are also much easier to manage on a daily basis. Cloning or snapping of workloads is done at a VM or VM-folder level.
For example, if an admin would like to create a testing and development environment, she can simply clone the set of virtual machines comprising that environment and users are ready to go. The outcome is massive reduction in time to provision test/dev environment – requiring only minutes instead of days.
One console to rule them all
With hyperconvergence putting both compute and storage under the same umbrella, there’s now an opportunity to create a much more intuitive process for allocating and managing resources – using a single management interface.
Springpath extends vCenter to include all relevant functionality for hyperconvergence – without the need to learn a dedicated management console for this platform. In fact, most of our customers report no real learning curve for daily operations. As you can see in the screenshot below, all cluster information is displayed in the Springpath tab, and operations like cloning, although offloaded to the Springpath Data Platform, can be done using the menus that admins are already using in their virtualized environment,
These familiar, integrated capabilities make a VM admin’s life simpler and more productive. But at some point there’s a need to scale – and Sprignpath can help big time in that area.
This will be the topic of our next post around A – Adaptive Scaling, so make sure to check back soon.