The time is nigh to address what should keep CIOs up-at-night: the zombies lurking in their data center.
As companies look to expand and upgrade their data centers, it’s easy to forget about the ghosts of servers past that are hiding in the background, continuing to consume electricity but serving no purpose. According to a study by NRDC, roughly 12 million U.S. servers in operation are doing little to no work but still drawing power. In fact, up to 30 percent of servers are “comatose,” meaning that they are no longer needed but still using significant amounts of energy, and many more still are grossly underutilized.
This threat is known but often overlooked as a server can easily turn from a ghost to zombie in a heartbeat. In addition to wasting energy and money, these unchecked zombie servers are vessels for malicious attack, proving the perfect under-the-radar location for external threats. Unless servers fail, administrators have no reason to check for abnormal energy use by individual servers.
IT teams need to take a proactive approach to identify and stop the growth of these silent threats. This will cut down on energy expenditures, reduce inefficiencies, and prevent harmful data breaches. In this slideshow, Muralidharan Palanisamy, CPO of AppViewX, a global leader in the management, automation and orchestration of data center services, walks us through best practices for preventing zombies from terrorizing your data center.
Stopping Data Center Zombies
Click through for tips organizations can use to prevent server zombies from terrorizing your data center, as identified by Muralidharan Palanisamy, CPO of AppViewX.
Make Room for Asset Management
Enterprises need to have a well-conceived and thorough asset management plan that takes into account the physical location of all servers, and helps determine how these servers play into the overall operation of the data center. This will help organizations track changes made to applications on these servers, providing a holistic view of their utilizations. If you don’t have a robust asset management system in place, IT teams can start building one by collecting information on any changes to the servers or its related components (e.g., firewall rules referencing the server). Over time, teams will be able to build an identity for these servers that can be tracked and referenced as their data center continues to expand and update. However, maintaining a record of the servers will only yield benefits when the documented data is used to preempt the emergence of zombie servers.
Use Networking Scanning Tools
Another way to get ahead of zombie servers is to use free network scanning tools to better understand the current scenario. Although not always comprehensive, because most machines have varied network security settings, it’s a step in the right direction and would help teams keep track of the ghosts of servers past lurking in their data centers, before they have an opportunity to become zombies. Over time, this would lead to electricity – and financial – savings that could be reassigned to support functional components of the data center.
Keep Track of Data Center Finances
Most finance teams keep detailed records of hardware and software purchases in their archives, and looking through them would help organizations arrive at an accurate and consolidated list of machines that can then be monitored. By knowing the MAC addresses and analyzing expenditure justification reports, IT teams can track even the most hidden servers and gauge whether or not they’re being utilized. Although somewhat tedious, keeping track of these records saves time and money in the long run.
Maintain Proper Procedures
Enterprises need to have considered IT procedures in place to ensure that servers don’t go comatose. IT teams need to demand justification for new server requests, keeping tabs on the details of their tentative lifecycle. These procedures should also be able to identify owners who would provide annual updates to perform a complete audit on necessary services, including cloud services, and physical and virtual machines. This would also save a lot of time and money when trying to identify and remove zombie servers.
Outsource to the Cloud
Zombie servers are more prominent in smaller data centers because their servers aren’t typically governed by a robust asset management system like their larger counterparts. That being said, smaller facilities should consider outsourcing their IT operations to the cloud. This option provides IT teams with the flexibility to scale faster and more quickly overcome errors from inadequate planning. This also helps data centers scale up, avoiding the purchasing of more physical hardware that may, over time, result in zombie servers.
