Configuration Management to Keep Virtualization On Target
With virtualization, data centers consolidate their hardware investments, and use their server hardware more efficiently, expanding application usage through virtual servers. Through consolidating, the number of physical servers can be greatly reduced. This alone brings benefits such as reduced floor space, power consumption and air conditioning costs. It lowers the Total Cost of Ownership (TCO) by increasing the efficiency of server resources and operational changes, as well as virtualization-specific features.
At system development sites, servers are often used inefficiently. When different physical servers are used by each business division's development team, the number of servers can easily increase. Conversely, when physical servers are shared by teams, reconfiguring development and test environments can be time and labor consuming.
Such issues can be resolved by using server virtualization to simultaneously have various operating system environments running on a single physical server, enabling parallel development and test of multiple environments.
However, the advantages of virtualization can be canceled out when the data center does not have an effective management strategy for the sprawl in virtualization configuration. This complexity can impact performance at many organizations where virtualization has been implemented.
Virtualized server deployment is continuing at the breakneck pace. InformationWeek reported that 54% of IT professionals expect more than half of their servers to be virtualized by 2011's end.
This growth of virtual environments makes management increasingly complex making it more difficult to optimize and control. Consider these areas of virtualization management to ensure that the benefits promised through virtualization are not lost.
Virtualization Complicates Capacity Management
A recent Virtualization Management Index study showed ( including over 1,400 environments, containing between 50 to 5,000 virtual machines a piece) that the broader the environment, the worse the density load is. In the physical environment, exactly the opposite occurs as the larger environment should have a better ability to balance and shift loads, while maximizing underlying resources.
The problem that arises is the poor management of virtual server capacity. Sys admins over-allocate system resources to compensate for another virtualization difficulty: a lack of visibility – impacting performance levels. Critical applications that lie at the core of the business should not be deployed to questionable environments. This allocation of additional resources as compensation for uncertainty adds even more configurations to the environment. Not only is ROI reduced, taking away from the original cost saving promises of virtualization, but risk grows.
Lack of Automated ManagementToday's virtual platforms lack tools to automate configuration management in virtual environments. Management tools are required for server teams to reduce workloads and to improve their capacity for ensuring a smooth running environment. The virtual environment may contain complex applications that spans dozens and dozens of servers with different levels of permissions and have an evolving data scheme. Since recovery to the snapshot is not always possible, it's not very easy to stop at this point and just rollback to the previous image, as you can lose critical business functionality and critical business data. Also a snapshot only includes the information and data that is required to roll back the status of a VM to a previous point-in-time. If something goes wrong after the snapshot is taken, you can effectively throw it away -- starting over where you were before the snapshot.
Integrating Management into Virtualization PlatformsStatus quo policies and procedures for IT infrastructure are significantly impacted with the implementation of the first virtual machine (VM) host. Yet many IT professionals still believe that they don't need to address their configuration management processes and technology for their virtual infrastructure. Virtualization platforms will need to integrate configuration management solutions to automate the performance of areas such as VM monitoring, lifecycle management, performance/capacity management and information sharing.
Risk of Outages to Virtual Environments
The speed of change facilitated by virtualization requires measurable and enforceable policies for configuration management and change management. With less than five minutes required to provision, configure or relocate a new virtual server, the window of opportunity for detecting misconfiguration is short.To prevent unscheduled outages the virtual infrastructure requires even more rigorous controls and configuration management capabilities. Unauthorized, undocumented and untested changes to consolidated virtual servers could have unplanned impact event causing outages to many parts of the business that can't be fixed with a simple rollback. This means that for virtualized environments, a configuration management process is vital to retain the benefits and cost savings of virtualization.
The virtual data center is subject to some unique risks. The biggest of these is from changes. The Virtual Server system and the self-service portals allow administrators to allocate virtual machines unheeded. With virtualization, systems or VMs can be provisioned in seconds. When change causes outages in a virtualized environment, the ripple effect is much greater than it is in a physical network.
A survey, from the recent VMworld event, found that a majority of administrators make changes to their VMs several times a day. This all points to some pretty significant risks from that ever present element: human error. These risks can't be addressed efficiently by adding more personnel since budgets for staff won't grow in sync to the amount of VMs provisioned. Rather they require Change management and Configuration management that is specialized for virtualization and automatically evaluates the virtual environment especially for simple, minute changes initiated by common "user-error" misconfigurations. For this, automated management capabilties that dive deep to the most granular level of a virtual system's configuration parameters, applying analytics to this data will allow IT teams to extract actionable information, to focus on what matters most in ensuring peak performance and availability.
While advancing the speed of integrating and implementing virtualization, automating virtualization management is a clear-cut directive to address conflicts, as well as building a data center that is dynamic, cost-effective and efficient.