In a recent survey conducted by Diskeeper Corporation, IT managers revealed that what really keeps them up at night is a dual problem: needing more data storage and answering the demands for cost cutting. Virtualization has been the accepted solution and it works, providing a primary barrier to virtualization performance is effectively handled. Since a company operates at the speed of its IT networks, faster I/O performance is constantly mandated. But as the IT stack of technologies build, the ability of the IT manager to maintain efficiency of system resources declines. Files written in pieces to both the host and VMs create massive waste of I/O resources, and “grow-as-needed” disks (a key tool of virtual operations), create “bloated” free space. Virtualization performance consequently decreases at the I/O throughput level at its deepest point of the platform, the disk.

Manual or fixed schedule solutions cannot produce virtualization performance gains without generating unacceptable losses in time and money or creating an up-again-down-again performance profile. The ideal components of a true solution are scalability, intelligent control, automatic operation and efficiency (increasing productivity with fewer resources).

One of the keys to solving this issue is to create awareness between the host and the virtual machines running on it. Resources needed to optimize each VM are not coordinated and will cut across virtualization performance. Not optimizing leads to a steady decline in I/O bandwidth and because so many operating systems are converged on physical systems, the results can cause a negative impact throughout the IT environment.

Managing virtualization performance is a commitment to achieving greater efficiency and economy. While some of the issues related to virtual platform performance and compatibilities are broadly known, there have been few real solutions.

Diskeeper Corporation research and development has produced V-locity® 2.0 virtual platform disk optimizer, a product that addresses all these points and generates significant increases in I/O speeds while also increasing overall efficiency. V-locity is grounded on four functions that separate it from any other would-be solution.

• The ability to coordinate resource usage and ensure V-locity operations are 100% invisible.
• The ability to ensure data is written contiguously to the disk the majority of the time while making the rest contiguous quickly with zero resource conflicts.
• Being able to detect virtual disk type and configure itself accordingly.
• Being able to compact virtual disk free space bloat on thin and dynamic disks.

The combined technologies in V-locity 2.0 are designed to ensure greater compatibilities between related functions and maximum I/O throughput across the entire platform. Virtual platforms at peak optimization exhibit economies of scale that go far beyond the sum of their parts. For example, a V-locity 2.0 user, the CTO of a large travel agency in charge of specialty and emerging markets (requires fast and flexible operations), operates a complex environment with over 350 servers. When his virtual systems were not performing up to expectations, he considered ordering higher performance fiber channel drives at $900 each, with a total cost of over $300,000. With V-locity deployment and a simple reconfiguration, he was able to get better response, consistently, with SATA drives that cost $50-$75 each.

Virtualization performance promises new levels of efficiency and flexibility that will continue to drive down costs for many years. IT managers who are recognizing the bottlenecks and taking advantage of the technology that resolves the issues will benefit sooner.