Java™ Scaling Feature Story

Scenario 3:

Breakthrough Scalability of an Application Constrained to an x86 Server

Problem: Too often, there is a mismatch between an application and the platform on which it runs. At Azul we recently looked at an application that consumes enormous resources on an x86 platform. A ‘big job’ could take over a week on a fast x86 system. While the application lent itself to parallelism, it was optimized to run on a small server. We tested the application to see what throughput could be achieved when hardware scalability constraints were removed and the application could access 384 cores and 128 GB of memory on an Azul compute appliance. Azul’s performance analysis and optimization tools were used to pinpoint bottlenecks and ease the process.

Solution: A whopping 45x improvements in scalability! The application was able to process 630,000 operations per second (OPS) compared to 14,000 OPS on the x86-based platform. The process to achieve this shines a spotlight on how traditional constraints limit scalability.

• Thread count: Initially the application was artificially limiting the number of threads that were allowed to execute. With few threads, the application started at 2,200 OPS on Azul (below the 14,000 native score) but by simply increasing the thread pool size and expanding the heap throughput jumped to 60,000 OPS.

• Memory locks: When many threads access the same piece of memory, traditional systems force developers to serialize the threads to make sure two threads do not simultaneously change the same memory location (similar to how airlines make sure the same seat is not assigned to two different people.) Azul appliances can detect such collisions in real time and assure correct execution. The removal of two such ”hot locks” and a further increase in the number of threads and heap size achieved a new peak of 115,000 OPS.

• Logging: Slow applications can afford to maintain logs of unneeded events or even multiple copies of the same information. As throughput increases, the shear volume of such logs make them difficult to analyze and it becomes more practical to log only what is needed. Reducing the amount of logging and addressing a related single threaded bottleneck raised the peak to 350,000 OPS.

• More locks: With the above steps taken, a new lock was exposed as a barrier to scalability. Once that was resolved the compute appliance was able to deliver 630,000 OPS – a 45x improvement over the original native performance!

The traditional solutions of scale up or scale out each present their own problems. Azul makes scaling simple and cost effective by providing large shared pools of capacity with the economics of commodity servers, eliminates the need to capacity plan at the application level, and makes adding capacity as simple as adding a new filer in a NAS environment. Beyond that the Azul architecture is ideally matched to the parallel multithreaded nature of Java™ applications. With enormous amounts of CPU and memory resources and optimization tools that help remove traditional constraints, developers can create new innovative applications that scale to levels impossible with traditional systems.

For more information on this study as well as how network attached processing can help your organization with similar transaction-intensive application deployment challenges, please contact sales@azulsystems.com or call 650.230.6650.