Intel Benchmarks Cape Clear 7 ESB Platform Capable of Processing 5 Million Complex Simultaneous Business Processes

San Mateo, Calif. - 20 March 2007

Sets New BPEL Reliability, Scalability, and Performance Standards for On-Demand Integration Projects

Cape Clear® Software, the Enterprise Service Bus (ESB) Platform leader, and Intel® announce that a benchmarking study designed to test the scalability and performance of Cape Clear 7's enhanced Business Process Execution Language (BPEL) engine has found Cape Clear 7 capable of processing 5 million complex and long-running simultaneously active business processes and 37 million transactions in a 24-hour period without compromising performance. With these results, Cape Clear 7 is setting a new performance standard for meeting the needs of mission-critical, on-demand integration projects currently in deployment across enterprise customers in the telecommunications, media, financial services, government, manufacturing, transportation, and energy industries.

On-demand integration, which is the fastest and most reliable way to integrate an organization’s systems across the internet and behind customer, supplier, partner and remote office firewalls using web services, requires supporting technology that guarantees optimal scalability and performance operating in a completely reliable environment. Cape Clear 7 delivers on this requirement with a productive developer toolset that significantly reduces ESB Platform implementation times.


Using industry-standard BPEL scenarios running on a network of up to eight CPU, Dual-Core Xeon® processors, Intel benchmarked the following key Cape Clear 7 results:

• Consistent transaction throughput of over 25,000 fully recoverable BPEL transactions per minute; equivalent to over 37 million in a single 24-hour period.
• Response rates actually improved as transactional load on the system increased.
• 5 million complex and long running simultaneously active instances were created; response times were shown to be constant and independent of the number of active instances.
• Full recovery was successfully demonstrated by randomly killing servers in the cluster and showing uninterrupted and correct processing behaviour while the processes were rapidly recovered.

Benchmarking BPEL presents unique challenges; for example, owing to the long-running and asynchronous nature of many of the applications, traditional Transaction Processing Council-style raw throughput measurements are not always appropriate. In support of this exercise, and based on the scalability and performance requirements of large-scale BPEL customers including Cisco, Channel 4, Pearson Education, MessageLabs, and Workday, Cape Clear has developed realistic and accurate scenarios to measure key performance, scalability, and reliability.

“The ability to provide server affinity for complex BPEL processes by horizontally scaling Cape Clear across servers while reducing response times is a significant achievement,” commented Tony Grewal, Project Manager at Intel. “This benchmarking study also demonstrates how off-the-shelf Intel-based hardware provides a powerful technical and economic solution for customers.”

“Large-scale BPEL deployments are a significant yet increasingly common challenge as enterprises deploy on-demand integration projects,” said David Clarke, senior vice president of products at Cape Clear. “These Intel benchmark results demonstrate that the Cape Clear 7 ESB Platform can easily meet the stringent uptime, reliability, scalability, and performance requirements of these large, mission-critical projects.”

The benchmark tests were run over a period of 10 days at the

Become a bobsguide member to access the following

1. Unrestricted access to bobsguide
2. Send a proposal request
3. Insights delivered daily to your inbox
4. Career development