Thursday, September 23, 2021

Table of Contents

Executive Summary

Background on This Paper

The History of TCO

TechWise Research's TCO Model

Who Was Surveyed

Company & Respondent Profile

Cluster Profile

Configurations Tested

The Value of an Hour

Possible Reasons for
Crashes and Downtime

Cluster Reliability Findings

Breakdown of the
TCO Components

Detailed Findings: Reliability-
Adjusted TCOTM for 2-Way Clusters

Detailed Findings: Reliability-
Adjusted TCOTM for 4-way Clusters

Detailed Findings: Reliability-
Adjusted TCOTM for 12 to 16-Way Clusters

Impact of Application Crashes
on the Cross-Over Point


Total Cost of Ownership for Entry-Level and Mid-Range Clusters

A Detailed Analysis of the Total Cost of Ownership of Three Different RISC-Based Server Clusters including HP OpenVMS, IBM AIX and Sun Solaris. (February 2004)

The overall concept of Total Cost of Ownership for server clusters has been an important issue within the IT community for many years. Numerous studies and customer experiences have proven that purchase price alone is not an adequate measurement to compare server clusters from various vendors. Other factors, including the costs to manage and maintain the servers, as well as the application availability they provide, usually have a greater financial impact on an organization than just the system's purchase price. Recognizing these factors, TechWise Research developed an analytical approach in 1999 called Reliability-Adjusted Total Cost of OwnershipTM that incorporates management costs and application availability in the TCO analysis. As part of this analysis, server clusters from different manufacturers are compared in terms of the actual number of downtime hours per year that customers typically experience. These downtime findings are then converted into a monetary measurement of the cost differences between clusters which TechWise Research refers to as the "Availability AdvantageTM."

In the past year, the issue of data security and application availability has moved from the realm of the IT manager into the corporate boardroom. Several high profile viruses and worms resulted in excessive Internet traffic bottlenecks and downtime during 2003. For this reason, TechWise Research expanded its proprietary Reliability-Adjusted TCOTM technique to incorporate downtime caused by software viruses and worms. To our knowledge, this is the first paper that analyzes three leading RISC-based clusters on this important issue.

This study focuses on the following three entry-level and mid-range server clusters: HP OpenVMS/ AlphaServer, IBM AIX/pSeries, and Sun Solaris/Sun Fire clusters. In the Fall of 2003, TechWise Research interviewed a total of 94 IT professionals in U.S. firms. The purpose of these interviews was to collect data on the operational costs associated with installing, managing, and maintaining their clusters. Information was also collected on the number of hours and associated costs for various downtime events each company experienced over a twelve-month period. All of these operational data were then combined with current system and service pricing (from IDEAS International) to calculate the Reliability-Adjusted TCO - an analytical approach that factors downtime costs/rates into the TCO analysis. TechWise Research included four main cost components in the TCO analysis. These are the costs to (1) buy the servers and service contract, (2) install and configure the cluster, (3) manage and maintain the cluster over three years, and, (4) the costs associated with application downtime over three years. Three different cluster configurations (2-way, 4-way and 12 to 16-way systems) were analyzed for the three platform brands. For each configuration, the Reliability-Adjusted TCO was calculated at various downtime costs to allow readers the ability to compare the different platforms at a downtime cost rate that most applies for their firm.

Study Results: For entry-level 2-way clusters, the acquisition costs (list price of the servers and service agreement) only represents 7% of the three-year TCO. As the clusters increase in complexity to a 4-way, then to a 12 to 16-way cluster, acquisition price represents a larger portion of the TCO, specifically 22% and 26%, respectively. However, regardless of configuration, the vast majority of the total cost of ownership is due to management and downtime costs. In terms of downtime, nearly half of the study's respondents report that their company loses as least $10,000 for each hour their cluster is down. The average cost per hour of downtime is $145,000.

For all three configurations (2, 4 and 12 to 16-way), the brand offering the lowest TCO changes depending on the cost per hour of downtime. In all three cases, when there are no costs associated with downtime, Sun Solaris/Sun Fire clusters offer the lowest TCO of the three brands tested. As the cost per hour of downtime increases, Sun loses its TCO lead to HP OpenVMS/AlphaServer clusters. The "cross-over" point, where Sun loses its "best in class" TCO status to HP, varies for the three configurations. For 2-way clusters, HP has the lowest TCO when downtime costs are more than $1,585 per hour. For 4-way clusters, HP offers the lowest TCO for downtime costs greater than $8,004 per hour. Lastly, for 12 to 16-way clusters, HP offers the lowest TCO when downtime costs are greater than $18,251 per hour. In all three configurations, IBM clusters' TCO falls between HP and Sun - with one exception. IBM has the highest TCO for 2-way clusters when downtime costs are less than $7,000 per hour.

Sun's TCO advantage at low hourly downtime rates can be attributed to the fact that Sun clusters have a lower list price than comparable clusters from HP or IBM. HP's TCO advantage at higher hourly downtime rates can be attributed to HP averaging the fewest hours of downtime for five of the seven causes of downtime tested. HP demonstrated a substantial Availability Advantage in the areas of crashes caused by software viruses/worms and end-user applications. For example, HP OpenVMS cluster owners reported an average of 0.88 hours of downtime per year due to software viruses or worms. Sun Solaris clusters were second best in this category averaging 4.32 hours, while IBM AIX clusters averaged 5.73 hours. For each $10,000 of hourly downtime costs, HP's Availability Advantage over Sun and IBM, for this one category, translates into $103,200 and $145,500, respectively, over a three-year period. In terms of operating systems, the OpenVMS operating system delivered the highest average availability of 99.990%, with IBM’s AIX coming in a close second at 99.987% and Sun at 99.972%.

When all seven types of crashes are considered, IBM AIX/pSeries clusters averaged 8.98 hours more downtime per year than HP OpenVMS/AlphaServer clusters. Furthermore, Sun Solaris/Sun Fire clusters averaged 21.50 hours more downtime per year than HP OpenVMS/AlphaServer clusters. For each $10,000 of hourly downtime costs, HP's total Availability AdvantageTM over IBM and Sun translates into $269,400 and $645,000 of savings, respectively, over a three-year period. Any IT manager who ignores availability in their purchase decision would be ignoring the most expensive cost component of the server cluster. This white paper summarizes the results of the study.

To receive a free copy of the full report, please provide the following information.  We will then send you an email with instructions how to obtain your report.

To prove you are not a bot, type the characters you see in this picture.

Request Paper     


© 1998-2014 TechWise Research, Inc. All rights reserved.
Page Last Updated: 07/19/2014