Power Systems Servers Top ITIC Reliability Survey
Illustration by Charles Williams/madeup.org
Server reliability is a crucial metric that impacts every aspect of a company’s daily operations and business outcomes. If servers fail and applications are unavailable, business stops. For the sixth year in a row, corporate enterprise users said IBM’s family of Power Systems* servers delivered the highest levels of inherent reliability and the least amount of outages among 14 server hardware and 11 different server virtualization platforms.
Cost of Downtime 98% said their average hourly downtime costs exceed $100,000
To ensure objectivity, Information Technology Intelligence Consulting (ITIC) accepted no vendor sponsorship in its independent ITIC 2014-2015 Global Server Hardware and Server OS Reliability Survey, which polled C-level executives and IT managers at more than 600 organizations worldwide last September. Top survey findings include:
- Fifty-eight percent of IBM Power Systems servers achieved five nines, or 99.999 percent availability. This is the equivalent of just 5.25 minutes of unplanned per server/per annum downtime, or 44 seconds each month.
- IBM Power Systems servers running IBM i and AIX* as well as the high-end System z* servers averaged the lowest percentage of annual server downtime. Only 4 percent experienced downtime of one to over four hours, compared to 6 percent of Hewlett-Packard servers and 8 percent of Oracle servers.
- Survey respondents also said IBM Power Systems servers recorded the least amount of prolonged outages of over four hours per server/per annum when compared to x86-based hardware platforms. Only 1 percent of survey respondents experienced prolonged outages on their Power Systems POWER7*, POWER7+* and POWER8* servers (which began shipping in June 2014) exceeding four hours.
Reliability’s Impact on the Bottom Line
These survey results support the prevailing industry trends centered on the need for higher reliability, uptime and security. Organizations demand high reliability to support virtualization and cloud environments as well as mobility, remote access and social media.
Because business is conducted around the clock across all global time zones, even a few minutes of downtime—particularly during peak usage hours—can result in significant costs and business losses. It can also cause internal and external business operations to grind to a halt. When the servers and the applications running on them are unavailable for any length of time, it has a domino effect that adversely and immediately impacts a company’s business capabilities. Reliability issues can damage a company’s reputation and cause it to lose business.
The latest ITIC 2014–2015 reliability poll indicates 79 percent of corporations now require a minimum of 99.99 percent uptime or better for their mission-critical hardware. This is a 12 percent increase from the 67 percent of respondents who said their businesses required four nines of reliability/availability in ITIC’s 2012-2013 survey.
In the mid-1990s, 99 percent uptime, which is nearly 88 hours of per server downtime, was acceptable. In the digital era of always-connected networks, 99 percent or even 99.9 percent availability is unacceptable and unthinkable. The accepted metrics of three or four nines of uptime equate to 8.76 hours and 52.56 minutes, respectively, of per server, per annum downtime.
comments powered by