August 8, 2018
For the tenth straight year, IBM and Lenovo servers again achieved top rankings in ITIC’s 2017 – 2018 Global Server Hardware and Server OS Reliability survey.
IBM’s Z Systems Enterprise server is in a class of its own. The IBM mainframe continues to exhibit peerless reliability besting all competitors. The Z recorded less than 10 seconds of unplanned per server downtime each month. Additionally less than one-half of one percent of all IBM Z customers reported unplanned outages that totaled greater than four (4) hours of system downtime in a single year.
Among mainstream servers, IBM Power Systems 7 and 8 and the Lenovo x86 X6 mission critical hardware consistently deliver the highest levels of reliability/uptime among 14 server hardware and 11 different mainstream server hardware virtualization platforms. Each platform averaged just 2.1 minutes of unplanned per annum/per server downtime (See Exhibit 1).
That makes the IBM Power Systems and Lenovo x 86 servers approximately 17 to 18 times more stable and available, than the least reliable distributions – the rival Oracle and HPE ProLiant servers.
Additionally, the latest ITIC survey results indicate just one percent of IBM Power Systems and Lenovo System x servers experienced over four (4) hours of unplanned annual downtime. This is the best showing among the 14 different server platforms surveyed.
ITIC’s 10th annual independent ITIC 2017 – 2018 Global Server Hardware and Server OS Reliability survey polled 800 organizations worldwide from August through December 2017. In order to obtain the most accurate and unbiased results, ITIC accepted no vendor sponsorship.
Survey Highlights
Among the other key survey highlights:
- IBM Z Systems Enterprise mainframe class servers topped all platforms for reliability, accessibility, performance, and security among. The Z servers deliver true mainframe fault tolerance experiencing just 0.91 – less than one minute of unplanned per server, per annum annual downtime. That equates to a barely perceptible 7.6 seconds per month or “blink and you miss it,” 1.8 seconds of weekly unplanned downtime.
- The z Systems family of servers also the lowest incidence – 0% — of > 4 hours of per server/per annum downtime of any hardware platform.
- To reiterate, IBM Power Systems and Lenovo System x servers once again consistently demonstrated the least amount of unplanned downtime 2.1 minutes per server/per year among all mainstream server offerings. Both the IBM and Lenovo offerings shaved several seconds off their previous results of 2.5 and 2.8 minutes per server/per year, respectively in ITIC’s 2017 Mid-year Reliability poll in June. These latest results underscore the stability and consistency of the IBM and Lenovo distributions running Linux as well as top notch service and support. Additionally, both vendors are aggressively updating their platforms to support Data Analytics, Artificial Intelligence (AI), Machine Learning and stronger embedded security and management capabilities.
- And 88% of IBM Power Systems and 87% of Lenovo System x users running RHEL, SUSE or Ubuntu Linux experience fewer than one unplanned outage per server/per year.
- Only one percent of IBM and Lenovo servers recorded over four (4) hours of unplanned per server/per annum downtime; followed by six percent of HPE servers; eight percent of Dell servers and 10% of Oracle servers (See Exhibit 2).
- IBM and Lenovo hardware and the Linux operating system distributions were either first or second in every reliability category, including virtualization and security.
- Fujitsu PRIMEQUEST and newcomer Huawei’s Kun Lun server platforms both scored high reliability rankings. It’s noteworthy that both the Fujitsu and Huawei server distributions are mainly deployed in Asia/Pacific and Europe. To date, neither has seen significant deployment in the North American region. However, that may change as competition in the server market intensifies. Huawei, in particularly is making a concerted effort to expand its global presence. Huawei continues to expand the performance, features and functions across its entire server line with an emphasis on its high end mission critical Kun Lun servers, which impressively tied for third place in reliability. Huawei has also redoubled its efforts to support red-hot cloud computing, Artificial Intelligence (AI) and Internet of Things (IoT) technologies and has inked a number of key software alliances and partnerships.
- Cisco UCS and HPE Integrity Superdome servers are getting more competitive with each subsequent ITIC Global Server Hardware and Server OS Reliability survey. In this latest 2017 Mid-Year Update poll, only two percent of Fujitsu PRIMEQUEST and three percent of Cisco, HPE Integrity and Huawei Kun Lun servers recorded in excess of four (4) hours of unplanned downtime.
- Lenovo x86 X6 mission critical servers scored the highest reliability ratings among all x86 platforms.
- Commodity Aging Hardware: Some 69% of survey respondents said aged hardware (3 ½+ years old) had a negative impact on server uptime and reliability vs. 31% that said it has not impacted reliability/uptime.
- Reliability continues to decline for the sixth year in a row on the HPE ProLiant and Oracle’s SPARC & x86 hardware and Solaris OS – each of which experienced 37 minutes and 33 minutes of unplanned outages per server/per annum. Reliability on the Oracle platforms declined slightly mainly due to aging. Many Oracle hardware customers continue to eschew or prolong upgrades and instead migrate to rival platforms.
Conclusions
In summary the ITIC 2017 – 2018 Global Server Hardware and Server OS Reliability Survey findings indicate that the reliability and uptime of the core infrastructure: server hardware, server operating systems and server virtualization are all more crucial than ever to the health of the overall on-premises datacenter and cloud-based environments.
Server hardware, operating systems, mission critical applications and networks must be available 24 x 7. Reliability and availability are crucial to uninterrupted business operations and productivity. IBM and Lenovo once again continued to deliver the highest levels of reliability. IBM’s z Systems Enterprise mainframe is in a class by itself and is the clear winner in terms of unparalleled fault tolerant levels of 99.999% or greater reliability.
Among the “work horse” mainstream systems, the IBM Power Systems, Lenovo System x are the most reliable and offer the highest levels of technical service, support and customer satisfaction.
Cisco UCS reliability is likewise very strong and consistent. The most recent ITIC survey also showed impressive gains by veteran vendor Fujitsu’s PRIMEQUEST E Series and newcomer Huawei’s Kun Lun mission critical server. To date, though these firms’ server offerings do not have a substantial presence in the North American region, though they are strong in Asia Pacific. Their strong showing in reliability makes them worth watching.
Organizations must have confidence in the reliability and stability of their server hardware and server OS platforms. The underlying advances in semiconductor, software and management technology bolsters the reliability of the servers. However, these technology gains can be undone by human error, increasingly complexity and the ever-present threats posed by Security, BYOD and IoT.