Nearly nine in ten IT managers believe virtual machine management is risky without proper planning, according to a global survey from Kelton Research commissioned by TeamQuest.
More than four in five IT managers who use virtual machines list these as top issues:
- Understanding what configurations will yield the best possible performance
- Assessing the performance and workload of virtualization machines
- Resolving performance problems easily
Other issues faced by IT include:
- Difficulty in comparing the cost effectiveness of different solutions
- Management believing the hype from virtualization vendors
- Evaluating the server sizes that are needed to support virtual containers
“One of the major problems with virtual machine management is that many organizations haven’t reached the point where they’re looking at latency instead of just utilization. Utilization doesn’t tell the real story, but latency and understanding components of response time does,” said TeamQuest Director of Product Management Scott Adams.
According to the study, companies using proper virtual machine management experience fewer cloud outages, too. Nearly 40 percent of respondents with proper virtual machine management have experienced cloud outages, compared to 66 percent of those without proper management.
Adams highlights the success of organizations using proper virtual machine management strategies. Such organizations spend less than half the time (18 hours/week) dealing with unexpected issues than their less mature counterparts (37 hours/week), according to the survey.
“These survey findings shout efficiency,” said Adams, “and as they increase their experience and maturity in these environments, they’ll start to realize significant value in proper planning. Things like knowing appropriate VM density, understanding how VMs interact, and determining the best configuration – that kind of planning leads to significant efficiency gains.”
Adams also suggests that organizations focus on solutions that provide comprehensive problem analysis capabilities across multiple vendors and technologies. “Today’s environment is too complex and layered, too diverse to try to make sense of the mounds of performance data from so many different sources. It’s much more efficient to use a solution that points you to the information that matters when problems arise,” said Adams.