1. Avoid the market hype and adhere to enterprise operational best practices.
Let's get serious. Sprawl is marketing speak hyperbole. No virtualized infrastructure apocalypse is threatening today's enterprise as no seasoned system administrator would permit rampant, ad hoc, and unsecured virtual machine proliferation to put an enterprise at risk. Sprawl is a byproduct of virtual machine proliferation that is representative of virtualized infrastructure growth. And, as has always been the case with datacenter technology innovation adoption, smart growth follows adherence to tried and true operational business practices.
2. Bottom line: Managing virtual environments is fundamentally different from managing physical ones.
While operational best practices are your friend, they must be applied differently to meet the reality of the virtualized infrastructure paradigm shift. As such, you will need to introduce new technologies and processes and be prepared to adapt.
3. You rely on automated management tools to maintain infrastructure visibility in the physical world; don't abandon this practice in the virtual world.
Virtual infrastructures are both transient and dynamic. If managed in an ad hoc fashion, it is impossible to maintain accurate visibility into and accountability for your virtual machines, thereby increasing corporate risk. Using Excel spreadsheets and other standard office tools in an attempt to manage virtual environments opens door to significant human error and does not scale. As is the case in the physical world, enterprises must efficiently tag, track, and report upon all virtual machines as they move throughout preproduction and production environments.
4. Granular Control can be both friend and a foe. Enterprises must follow an intelligent, controlled approach to growing virtual machine instances.
Virtualization offers enterprises unprecedented opportunities to increase agility, reduce costs and operate more efficiently. As virtual machines become pervasive, users will be able to create, modify, and distribute new "machines" with unprecedented ease. This flexibility, while providing tremendous benefit, can also undermine many of today's risk management approaches, with negated cost savings and reduced agility the result.
5. Seek out management and security tools purpose-built for virtualized environments.
Virtualization in the datacenter typically leads administrators to load up everything they used in the physical world to compensate for their virtual machine loss of control. Monitoring agents, configuration management applications, and commoditized security technologies are applied. The simple result: Virtualization scale benefits are lost and performance suffers. Utilizing technology purpose built for virtualization's requirements for efficiency, a small footprint, and portability in a dynamic virtual environment is key.
6. Rethink your network defenses.
Most network defenses are predicated on seeing traffic. However the definition of the "network" in a virtualized world is significantly different. Today's virtual environments require the monitoring of inter-process communications within virtual machines and across multiple physical machines.
7. The virtualization freight train is coming and new best practices are on the horizon.
The virtualization industry will continue to create operational deployment and security best practices. Enterprises should create a centralized policy store of industry best practices for enforcing virtual machine operational policies running on desktops, workstations, servers and hypervisors.
John Suit is CTO at Fortisphere.