Avoiding the VM zombies

Charles Clarke, Veeam technical director for APAC, explains how resellers can educate customers on minimising VM sprawl.

Many organisations believe virtual machine (VM) sprawl isn't an issue for their infrastructure, but the dynamic nature of virtualisation can create an army of walking dead VM's.

Resellers can help companies become aware of the potential risks arising as a result of 'virtual zombies', and educate them on how to avoid them.

VM sprawl is a symptom of any rapidly growing virtual infrastructure and can happen when organisations set up VMs for temporary usage, such as for projects, testing or development labs and then forget about their existence after peak usage. Virtual machines can outlive their usefulness, but be left behind as administrators are unsure of whether to delete them. This causes the VMs to essentially become 'virtual junk' within the organisation's infrastructure.

The chaotic increase of unmanaged VMs drives extra resource consumption. Even inactive VMs still utilise host storage and CPU, wasting resources that could be allocated to more heavily used applications. Eventually, VM sprawl will exhaust the environment. It can delay new projects and entangle budgets on extra hardware that could be spent on initiatives to advance the business.

Resellers should offer services and solutions that can detect key VM sprawl threats and provide analysis and recommendations on right-sizing and efficient allocation of resources. The best solutions will provide advanced capacity planning, which help to forecast, plan and oversee virtual infrastructure resources.

Beyond offering software, it is important to educate customers on how to avoid the problem.

Tips for virtualisation sprawl control 

Identify 'Zombies' Zombies are low-activity, idle VMs that are running, yet effectively doing nothing.

Zombie VMs can be identified through reports showing idle VMs identified through network communication or resource usage for example. Organisations can then decide whether to shut them down, reconfigure roles, decrease hardware provisioning or delete them altogether.

Eliminate excess backups Backup requires resources including storage, and can be a lurking space for zombies. There may be duplicate VMs included in several discrete backup jobs, sprawling within the backup repository. Zombie VM's still running in production may be backed up needlessly, consuming resources. Organisations should use tools that give insight into  backup infrastructure. Companies should also identify junk and keep the backups lean and efficient.

Clean up garbage Garbage files are the side effect of everyday changes in the organisation's virtual infrastructure. VM configuration data and temporary files may still remain on the data store and consume space after parent objects are deleted. Valuable VM storage can be used for non-VM data such as disk images and downloaded files. Enterprises should identify non-critical data and clean it up.

Categorise VMs Organisations should categorise objects in the virtual infrastructure. Each VM should be assigned to one or several groups by a variety of custom parameters, such as department, purpose and specific project. The organisation can then get an at-a-glance view of the virtual machines it cares about in that context. This helps spot zombies and make good management decisions.

Detect rogue snapshots If a snapshot falls out of the relevant snapshot-chain, it is an orphaned snapshot. This may occur after errors such as host failures, badly performed backups and unsuccessful-snapshot consolidation. Forgotten snapshots may consume significant storage. Organisations should use tools that manage snapshots well, and they should have a mechanism to hunt and delete old and orphaned snapshots.