A survey of more than 500 senior IT professionals reveals that a whopping 89 percent test their disaster recovery/failover systems only once a year or not at all, leaving their enterprises vulnerable to massive technology and business failures in the event of a disaster.
In addition, 70 percent of respondents said it would take at least four hours for their servers to recover completely from failure, including restored software, configuration and network and storage connectivity. More than 50 percent of those polled estimated it would take several days to bring their systems back up to full capacity.
The survey, conducted in March and April by researcher Brilliant Ideas LLC for Scalent Systems, also said that nearly 67 percent of respondents were only “minimally confident” that their disaster recovery system would work as planned in the event of a disaster.
These findings emphasize the business and financial risks faced by companies that have not taken steps to ensure they can quickly recover from IT disasters, a spokesperson for the San Jose, Calif.-based researcher said.
Data center downtime has been proven costly for companies. Industry researchers have estimated that a company loses an average of $84,000 to $90,000 for every hour of downtime, Scalent said. More than 90 percent of companies that experience one week of data center downtime go out of business within 12 months, according to the National Archives and Records Administration.
“Companies every day face a variety of threats that could bring their systems down,” said John Humphreys, a program director at IDC Research in Framingham, Mass.
“Even companies that dont rely on their IT infrastructure for high-volume business transactions should recognize that slow recovery in the event of a disaster can seriously impact customers, employees and shareholder value. A highly reliable, cost effective and straightforward DR and HA environment is quickly becoming a key to deliver business services that span server, storage and network elements.”
Poor failover performance and non-optimal DR strategies are considered two of the most important problems that put systems at risk, the report said.
“With so much riding on their IT infrastructure, companies should adopt a disaster-recovery approach that can recover large numbers of servers and their associated network and storage connectivity in less than five minutes, guarantee zero differences between production and back-up systems, and automatically validate DR capabilities on a regular basis,” said Scalent vice president of products Kevin Epstein in Palo Alto, Calif.
Scalent provides adaptive infrastructure software that encapsulates each existing servers software stack, placing it on central storage, and then virtualizes the network and storage connectivity of each physical server, so that any server or set of servers can appear to be any others on the network, Epstein said.
In the Scalent system, if any physical server fails, any other server can be “turned into” the failed server in the time it takes to reboot. This “rack once, cable once” scenario means that there is no need for manual re-imaging or re-cabling, and the replacement server is an exact duplicate of the original, eliminating configuration issues.
In addition, Scalent software does not degrade server performance, nor does it introduce a new point of failure, because it does not sit in the data path, Epstein said. Should Scalents controller ever fail, servers simply maintain their existing topography until Scalent is restarted. There is no service interruption, Epstein said.
Other Companies also Touting Adaptive Infrastructure-Type Products
The companys main rival in this space, Hewlett Packard, employs an entire division to provide an adaptive infrastructure product line for its server and storage server systems.
Cisco Systems will talk to anybody wholl listen about how it now wants to be known as the “business intelligence” within the adapative infrastructure of the network—to improve both security and reliability.
For years, IBM has been talking about autonomic computing, which is similar to adaptive infrastructure in that it includes business intelligence that gives computers self-healing and load-balancing qualities in order to withstand outages and other problems that affect daily performance.
Intel has worked closely with IBM to provide processors optimized to do the virtualization that is at the heart of these high-performance computing models.
Scalent Extends Support to IBMs AIX Mainframe System
In other news, Scalent on May 2 announced it will extend its frontline V/OE software to support the IBM AIX mainframe operating system and architecture.
Scalent will certify IBMs System p line of servers, extending its current supported processing architectures—Windows, Linux and Unix. Scalent will provide an adaptive infrastructure software solution for AIX, enabling AIX customers to increase data center flexibility and performance for business continuity and automation initiatives.