Take Five With Tom Fenton

Five VM-Level Infrastructure Adaptations

Infrastructure is evolving for the better, making the job of the admin easier in the long run. Here are five ways it's evolving to work at the VM level.

• By Mike Matchett
• 07/21/2014

It used to be that IT struggled to intimately understand every app in order to provide the right supporting infrastructure. Today, server virtualization makes the job much easier, because IT can now just cater to VMs. By working and communicating at the VM level, both app owners and infrastructure admins stay focused, using a common API to help ensure apps are hosted effectively and IT runs efficiently.

But the virtual admin still has to translate what each VM requires, going beyond direct-server resources into the specialized domains of other IT infrastructure silos. While silos have traditionally pooled rare expertise to optimize expensive resources, in today's virtualized world, silos seem to offer more friction than leverage. Here are five ways infrastructure is evolving to work at the VM level.

TAKE 1
VM-Centric Storage. VMware Virtual SAN and external storage arrays such as those from Tintri are already delivering storage services at the VM level, to make aligning per-VM storage requirements with actual storage resources automatic and optimized. VMware Virtual Volumes (VVOLs) promises to extend this VM-centric approach to many expected third-party storage arrays. These arrays will need to deliver per-VM storage services, rather than "hands-off" abstract storage silo capacities like the traditional LUN or an aggregate datastore.

TAKE 2
Data Protection and Disaster Recovery. Backing up at the increasingly obsolete LUN or datastore (which can contain multiple VMs, or each VM can have multiple storage units) is giving way to being able to back up at the VM level directly. Backups, snapshots, clones, and replication at the VM level (with provision for application-level consistency) mean that applications can independently specify protection and recovery requirements. In the best solutions, policies can be set per-VM to direct and inform downstream infrastructure, ensuring appropriate protection without separate management.

TAKE 3
Provisioning. Of course, clouds are based on the elastic provisioning of instances, but we can now look for an emerging ability to seamlessly migrate VMs (and all their associated resources) across multiple clouds. By working at the VM level (an application here might mean a defined group of related VMs), IT should be able to directly manage and optimize an organization's leveraged use of public, private, and increasingly, hybrid cloud infrastructures.

TAKE4
Convergence and Hyperconvergence. Many infrastructure vendors (such as HP, IBM, VCE) are converging, or even hyperconverging (Simplivity, Scale Computing, Nutanix) server, storage and networking resources into unified platforms. In a converged solution, the basic broad unit of the converged resources is essentially the VM. These systems tend to be specified and deployed to support identified collections of VMs as a whole system.

TAKE 5
VM-Level Management. If the VM is the main resource that IT manages from both upstream to apps and downstream to infrastructure, then it's key for IT to have system management capabilities that start with the VM. This isn't easy in a dynamic, virtualized, cloud-hosted, software-defined world, yet continues to be essential. Underneath it all, physical resources will still fail or bottleneck, so look to management solutions that map from physical to virtual, across on-premises to cloud, and maintain currency automatically as systems dynamically change.

If the VM has become the de facto IT application, then the VM is also the new "resource" infrastructure object. In the new software-defined datacenter world, we should expect to soon see all types of dynamic infrastructure providing service in VM terms.