IBM Systems Magazine, Power Systems - November 2017 - 22
IBM Power Systems comes out on top in latest ITIC global reliability survey
By Laura DiDio
he saying, "Time is money"
perfectly encapsulates the
reality of daily operations
in today's complex and demanding corporate IT environments.
Reliability and continuous,
uninterrupted access to network
resources comprise the essential
foundation for daily business
operations. High availability (HA)
is crucial in the 21st century
where servers, OSes, desktops,
mobile devices and applications
are increasingly interconnected
via the Internet of Things (IoT),
virtualization and the cloud.
When data is inaccessible for any
reason, business ceases. This has
a domino effect on the corporation
and its end users, customers, business partners and suppliers.
According to Information Technology Intelligence Consulting's (ITIC)
2017-2018 Global Server Hardware and Server OS
Reliability survey, 80 percent of businesses surveyed require a minimum of 99.99 percent uptime-commonly
referred to as "four nines." That's double the 39 percent
of respondents who indicated they required four nines
of uptime five years ago. The independent web-based
survey polled 800 executives and IT managers at corporations worldwide in August and September. In order
to obtain the most accurate and objective results, ITIC
accepted no vendor sponsorship.
Additionally, 17 percent of respondents indicated
their businesses now require "five nines" or 99.999
percent server and operating system uptime. That
equates to 5.26 minutes a year, 25.9 seconds a month
or 6.05 seconds a week of unplanned downtime per
server (see Figure 1, page 24). Moreover, 3 percent of
leading-edge businesses-mostly in highly regulated
industries such as banking, finance, government
and healthcare-need 99.9999 percent uptime. This
near-flawless server availability equates to 31.5 seconds of unplanned downtime per server, per month.
"Reliability is paramount. There's zero tolerance for
downtime here," says Jim Brown, a software quality
assurance analyst and system administrator at Boston
University. "If BU students and productivity workers
22 // NOVEMBER 2017 ibmsystemsmag.com
can't access their data and applications for even a few minutes, it's
problematic for all concerned."
Brown says BU is very proactive
in both quality assurance (QA) and
reliability testing of all servers and
software, while also prioritizing vendor technical service and support.
"We need a minimum of four nines
and preferably five nines of availability. But invariably, situations can
arise that cause unplanned server,
OS and application downtime.
When that happens, we need to
know that our vendors will be right
there to assist us," Brown says.
Andrew Baker, CIO at Brainwave
Consulting, a Charleston, West
Virginia-based security consultancy, agrees. "The only good
downtime is no downtime," Baker
says. "Once a server goes down
or an application is unavailable,
companies only want to know one
thing: 'How fast can you get us up