Virtual Machine Sprawl
Server virtualization has enabled IT administrators speed, flexibility, and ease in providing platforms and workloads to their end customers with unprecedented service delivery times. This speed of delivery has enabled many IT teams to transform and remain agile to the business systems that they support. Unfortunately, because of the ease in delivery, the rampant perception by business owners is that virtual machines are free because they have no physical presence and therefore requests for new virtual machines (VMs) typically meet little resistance. This problem has become so common with in IT organizations, that it is often referred to as “VM Sprawl”. With VM Sprawl the number of VMs running in a virtualized infrastructure increases over time simply because of the ease of creating new VMs, not because those VMs are absolutely necessary for the business. The reality is that virtual machines are NOT free and they have the same licensing, support, security and compliance issues that physical machines do. To prevent “VM Sprawl” IT professionals should define and enforce a process for VM deployment and periodically understand when under-utilized and dormant systems can be retired and archived.
DataGravity Discovery Series – VM Integration and Protection
The DataGravity Discovery Series enables IT professionals to thoroughly maximize their investment in server virtualization, while empowering them with capabilities to enforce a process for managing an efficient VM life-cycle. The system was designed to provide search, governance, activity monitoring and protection capabilities on a per-VM basis, extending its capabilities not only on the VM container, but also into the VM file system. This level of granularity is paramount in the effort to reduce VM Sprawl.
Virtual machine integration allows a holistic view of the status of each of your VMs. This includes showing a virtual machine’s state, not only if it is running or powered off but also highlighting if it has been deleted on the VM datastore. In fact, most native VM management systems don’t have the ability to report on orphaned VMs, where someone accidentally removed the VM from the inventory but did not remove its files from the underlying datastore storage. This sometimes is done intentionally where a virtualization admin doesn’t want to have to manage the VM in the inventory, but does want to keep the files. Unfortunately there isn’t an easy way to see all VMs that exist on datastores but aren’t part of the inventory. This simply adds to the problem of VM management and VM Sprawl.
In order to reduce the amount of overhead involved with VM Sprawl, the DataGravity Discovery Series provides native reporting on the VM status on a per-VM basis, and offers protection and recovery capabilities allowing administrators to combat VM Sprawl. From the Manage VMs screen within DataGravity an administrator can quickly understand which VMs are consuming storage resources and how many of these resources are being consumed. The administrator can further filter on the status of these VMs – are they active and running, powered off, orphaned from the VM inventory, or deleted. It is easy for an administrator to focus in on VMs that can be cleaned up.
DataGravity provides administrators with elegant retirement options for VMs that no longer need to be managed. Because of it’s integration with VM infrastructure and it’s inherent recovery services, an administrator has the ability to retire virtual machines from the infrastructure while providing recovery options should those VMs need to be added back to the VM inventory.
Leveraging the use of DataGravity’s DiscoveryPoints, administrators can safely retire VMs from their infrastructure knowing that restoration is only several clicks away. DataGravity DiscoveryPoints fully integrate with the VM infrastructure, are maintained on a fault isolated disk pool, and provide unprecedented recovery speed. To begin the retirement of a Virtual Machine, a permanent DiscoveryPoint is invoked on the VM from within the DataGravity Manage VMs screen.
Once the DiscoveryPoint has been successfully created, with a full indexed and cataloged set of VM files for that VM, it can then be powered off and removed from the Virtual Machine infrastructure.
Retiring Virtual Machines and Recovering Infrastructure Assets
To complete the retirement of a VM it can then be powered off and removed from inventory by selecting the virtual machine from within vCenter, performing a Power Off, followed by a Delete from Disk. This will ensure that the VM is no longer using compute and storage resources within the environment.
VMs that have been deleted from inventory are shown in the DataGravity Managed VMs screen with a status of ‘Deleted’.
Orphaned Virtual Machines consuming Storage Resources
Orphaned VMs, are those VMs that have been removed from the inventory but have not been deleted from the datastore, and are therefore still consuming storage resources. It is not uncommon to find orphaned VMs within an environment for several reasons: i.) someone accidentally removed the VM from inventory, ii) someone “deleted” the VM but actually only removed it, or iii) someone wanted the VM out of the inventory, but wanted to keep the files. Within vCenter there isn’t any report or easy way to see VMs that have been orphaned, which results in wasted storage resources otherwise unknown to administrators.
DataGravity Manage VMs allows an administrator to quickly find orphaned VMs, by listing them with a status of ‘Unknown’, giving that administrator and opportunity to reclaim otherwise wasted storage space.
Virtual Machine Restores
In the event that a server was retired or deleted pre-maturely, they are extremely easy to recover. In fact an administrator can easily recovery a VM back to its pre-retired state in only several clicks. To recover a deleted or retired VM utilizing DataGravity DiscoveryPoints, is simply a matter of finding the VM on the Manage VMs Screen, selecting the appropriate DiscoveryPoint to restore to, and performing the restore. This process is outlined in 'Restoring a VM' as part of the Virtualized Recovery post that was written last week.
This fast proven plan for recourse with the increased visibility on a per-VM basis provided by DataGravity ensures confidence and provides administrators the tools they need to identify orphaned VMs and combat VM Sprawl.