Sujoy Saraswati

Compiler aided selective lock assignment for improving the performance of software transactional memory

Atomic sections have been recently introduced as a language construct to improve the programmability of concurrent software. They simplify programming by not requiring the explicit specification of locks for shared data. Typically atomic sections are supported in software either through the use of optimistic concurrency by using transactional memory or through the use of pessimistic concurrency using compiler-assigned locks. As a software transactional memory (STM) system does not take advantage of the specific memory access patterns of an application it often suffers from false conflicts and high validation overheads. On the other hand, the compiler usually ends up assigning coarse grain locks as it relies on whole program points-to analysis which is conservative by nature. This adversely affects performance by limiting concurrency. In order to mitigate the disadvantages associated with STMs lock assignment scheme, we propose a hybrid approach which combines STMs lock assignment with a compiler aided selective lock assignment scheme (referred to as SCLA-STM). SCLA-STM overcomes the inefficiencies associated with a purely compile-time lock assignment approach by (i) using the underlying STM for shared variables where only a conservative analysis is possible by the compiler (e.g., in the presence of may-alias points to information) and (ii) being selective about the shared data chosen for the compiler-aided lock assignment. We describe our prototype SCLA-STM scheme implemented in the hp-ux IA-64 C/C++ compiler, using TL2 as our STM implementation. We show that SCLA-STM improves application performance for certain STAMP benchmarks from 1.68% to 37.13%.

Steal-A-GC: Framework to Trigger GC during Idle Periods in Distributed Systems

Distributed applications for data analytics have emerged to be a significant growth domain both in terms of technology as well as business relevance. With the advent of Hadoop MapReduce, several distributed execution frameworks have been developed for data analytics and machine learning applications. Many of these frameworks are written in Java, Scala or Python and run on top of the Java Virtual Machine (JVM), leaving them vulnerable to the performance impact of garbage collection (GC) operations. Application pauses induced by the Java GC are notorious in being one of the most important contributors of runtime performance bottleneck for a single process execution. These pauses are present in the distributed scenario as well, though in this case, it is significantly more complex to measure the impact of a GC pause for the overall execution runtime. Typically, in a distributed workload, a GC can result in delayed tasks impacting the overall completion time of the application. We present herein a novel way to preemptively trigger Java GC, stealing idle cycles, so as to prevent future tasks from being impacted by a GC pause during peak execution periods, resulting in improved overall application throughput.

Performance Analysis of Java Virtual Machine for Machine Learning Workloads using Apache Spark

Now a days data is growing very rapidly, where processing and analyzing data to get useful information is the main task. There are many big data processing tools and framework such as Hadoop, Hive, Cassandra etc. Spark is one of the fastest big data processing framework in cluster computation. Basic Idea is to analyze the performance of java virtual machine (JVM) [1], by characterizing java virtual machine using SparkBench benchmark on Apache Spark™ [2]. Java virtual machine is a core execution platform for spark application. When we run the spark application on java virtual machine, its behavior is affected, which needs to be monitored to analyze the JVM performance. Here we are considering Machine Learning workloads like K-Means, Matrix Factorization and Logistic Regression. Main goal here is to analyze the machine learning workloads end to end across the cluster, with respect to following parameters such as garbage collection, memory such as heap usage, CPU process time. Characterization of JVM is done with spark cluster setup and HDFS is used as storage with distributed Hadoop cluster setup.

JVM characterization framework for workload generated as per machine learning benckmark and spark framework

Today there are plenty of frameworks to assist the development of Big-data applications. Computation and Storage are two major activities in these applications. Spark framework has replaced Map-Reduce in Hadoop, which is the preferred analytics engine for Big-data applications. Java Virtual Machine (JVM) is used as execution platform irrespective of which framework is used for development. In the production environment it is essential to monitor the health of application to gain better performance. The parameters like memory usage, CPU utilization and frequency of Garbage Collection etc., will help to decide on the health of application. In this paper a framework is proposed to characterize the JVM behavior to monitor the health of application. Workload generated by running Machine Learning algorithms available in Spark Benchmark Suite.