CiRBA 4.6 Raises the Bar on Workload Analysis through Advanced Benchmarking & Probability Analysis

By VSM News Staff
published: Tuesday, April 22 2008

CiRBA Version 4.6 Opens Up Cross-Platform Virtualization Opportunities and Greater Control Over Risk in Virtualized Infrastructure.

Toronto, Ontario - April 22, 2008 - CiRBA Inc., a leader in Data Center Intelligence software, today announced the release of Version 4.6 of its leading virtualization and consolidation analysis software.  With Version 4.6, CiRBA adds to its already unmatched workload analysis, enabling organizations to accurately plan and design virtualized infrastructure on any platform - including mainframes - according to workload personalities, risk tolerance, service level agreements and performance requirements. This advanced workload modeling also can consider the impact of iSCSI storage and network connectivity on virtualized servers to optimize placements and use of these technologies without adversely impacting performance.

Advanced Workload Benchmarking

Each application has its own distinct workload personality that impacts how it will perform on any given virtualization platform and how well it fits with other resources on a server. Benchmarking is critical in determining how applications will perform in a given environment in order to choose the best virtualization strategy and targets.  It is imperative that the right benchmarks are chosen considering the nature of the individual workload to be virtualized. CiRBA Version 4.6 helps teams determine which benchmarks are appropriate for specific workloads and then enables analysis using multiple benchmarks within a single virtual pool, resulting in accurate modeling of combined workloads. CiRBA supports widely used benchmarks including those from Standard Performance Evaluation Corp. (SPEC), Ideas International, and others.

Workload Contention Analysis

Risk mitigation is a major concern for organizations when consolidating production environments. Once it is determined that multiple servers are suitable for consolidation from a business and configuration perspective, workload analysis becomes pivotal.  CiRBA Version 4.6 provides probability analysis on workload contention to help drive decisions surrounding the level of risk the organization is willing to tolerate. For example, critical applications that must always provide high performance for transaction processing may be analyzed by specifying a very low "peak contention" tolerance, thereby greatly reducing the risk that workloads will compete for resources and performance will be adversely affected. In other scenarios, such as batch processing or non-production environments where performance is not the main concern, higher contention tolerance levels may be appropriate as the focus is on completion of jobs  rather than the speed with which they are completed. CiRBA provides a variety of workload scoring strategies based on peak or sustained activity to be applied in conjunction with tolerance ratings so that organizations can balance service levels, potential risk, and consolidation ratios.

This type of analysis also enables teams to establish pools of virtualized resources based on performance requirements and risk tolerance associated with specific applications. High and low priority applications can also be analyzed into the same pools to provide a balance of performance and overall utilization.  This benefits IT and business units alike as appropriate service levels and costs can then be negotiated, defined, and applied.

Analysis for iSCSI-based Storage within Virtualized Infrastructure

iSCSI SAN Storage is rapidly gaining acceptance as the low cost option for enabling centralized storage, particularly within x86-based environments. However, the increased network load this creates can impact networking interfaces and switches, requiring careful consideration of aggregate utilization levels.  Also, the use of non-offloading NICs can cause a tax on CPU utilization as the host constructs SCSI frames at the software level.  To properly assess the impact of this, CiRBA enables organizations to model iSCSI workloads from a bandwidth and CPU overhead perspective in order to determine how and where the technology can be leveraged effectively within the environment without adversely impacting performance.

Network Latency Analysis

Virtualization and consolidation may introduce changes in the communication topology of an environment. CiRBA Version 4.6 provides features to track the interconnect between servers as well as the latency patterns of the environment on the whole.  This allows analysts to understand what impact any changes will have on application communication paths, and highlights any transformations that may subject applications to increased network latency.  By ensuring that any changes are "transparent" from a network level, this feature complements CiRBA's server-level analysis to ensure that critical data center transformations are as risk-free as possible. 

Inter-Platform Consolidation Analysis Including Mainframes

As virtualization matures and the benefits of smaller consolidation initiatives are realized, organizations are becoming more sophisticated in how they view the opportunities provided by these technologies. Increasingly, organizations wish to consider which operating environments provide the greatest savings without compromising risk and performance levels for a given application. CiRBA Version 4.6 provides sophisticated rules-based analysis of business and technical constraints and workload analysis that enables organizations to determine which applications currently existing within UNIX and x86-based environments are suited to operate on mainframes. CiRBA accurately models how applications will perform in mainframe environments with its advanced workload analysis that includes new Version 4.6 functionality of advanced benchmarking, workload contention probabilities and network latency analysis.

"We see a significant demand for cross-platform transformations, including the migration of applications to mainframe environments," said Andrew Hillier, CiRBA Co-Founder and CTO.  "This requires cross-platform data acquisition, cross-strategy analytics and the ability to model the performance of different types of workload personalities on different platforms.  CiRBA 4.6 adds key features in this area, as well as advanced network and storage analysis, to provide a data center wide intelligence platform to transform and manage both physical and virtual environments."

About CiRBA Inc.

CiRBA is a leading provider of Data Center Intelligence (DCI) solutions. CiRBA provides enterprise software that enables Global 5000 organizations to rapidly analyze and visually map the safest path to an optimally consolidated and virtualized data center. Using a unique cross-platform, multi-strategy approach, CiRBA addresses all configuration, workload and business constraints to ensure success in any consolidation initiative. For more information, visit www.cirba.com.