OKLAHOMA CITY — A complete virtualization of its computing infrastructure has allowed Tinker FCU to dramatically reduce overnight processing and backup times, improve quality assurance and boost its disaster recovery efforts.
Though it took the $1.9-billion institution 18 months to go from starting the discussion to implementation, the project itself, which included virtualizing the Episys core platform to run on a single IBM P570 instead of four separate servers, installing a new server at the contingency site, and modifying its storage system to utilized a SAN design took only a few days. The hardware and software involved required an investment of only $157,000.
Thanks to the nature of a virtualized system, Tinker was able to use its disaster site as "guinea pig" for the conversion.
"When everything worked as designed we felt comfortable flipping the switch and putting it onto our active site," said senior vice president/CIO Grant Woldum.
With the switch flipped, the virtualization in place and the new backup system up and running, Tinker FCU has seen huge changes in a number of its operations. Overnight processing times have "shrunk drastically" from an average of four or five hours to just 45 minutes.
"That becomes even more significant when you do month-ins, quarter-ins and year-ins," VP-IS Operations Steve Mooney pointed out.
Symitar puts out new releases for its core system three times a year; under the old system thee implementation was quite a production.
"Before we had virtualization this was tested on another system independently of the core system," Mooney explained, noting that the test would take several days before IT staff even began the process of implementing the release on the live environment. With virtualization the test is done on the remote site and once completed, the release is instantly implemented on the live site.
"You're talking about a three to four day process shrunk down now to thirty minutes," Mooney added.
The change to a virtual system is also good for the environment and the bottom line; the credit union no longer has to buy a large amount of hardware to expand computing capacity saving on those potential purchases and the electric bills they would run up. The new backup system has eliminated the need for tapes and the couriers that would bring those backup tapes to the disaster recovery site, saving Tinker a bundle there. A job scheduler and a streamlining of processing has also enabled the credit union to eliminate four nightly operations positions.
Members are benefiting from the switch in tangible ways as well. The greatly improved processing system has enabled Tinker FCU to expand its call center hours and Woldum is confident the credit union will eventually be able to serve its members over the phone 24/7 because it is no longer constrained by lengthy overnight processing delays.
Best Practices CUJ 2009
CU: Tinker FCU
Best Practice: Virtualization of Computer Infrastructure
Vendor: Internal, Symitar
