High Performing and Predictable Java network programming

Title: High Performing and Predictable Java network programming

Speakers: Dinesh Kumar, Harshad Sane, Rodolfo De Vega, and Kshitij Doshi (Intel)

Abstract: Modern applications are increasingly structured as loosely coupled collaborations, working in networked interactions while producing and retrieving data from massive datasets through random I/Os. Solution developers, once used to the simplicity of doing I/O through synchronous calls, have to increasingly architect for asynchronous-streaming interactions among loosely coupled peers and data sources/sinks. It is common for I/O operations to not only be multiplexed at platform software level, but also to pass through multiple layers of protocol stacks and runtimes, and a variety of proxying-caching-transformation and filtering operations, before being presented to the core logic of a service. Transitions through these intermediary functionalities need to preserve non-blocking I/O interactions that are designed in at the lowest plane, e.g., at a user-kernel/application-runtime boundary. For programmers of core business logic in an object oriented or functional language such as Java, Scala, Kotlin, these benefits are realized through the Netty framework which lets programmers follow a keep-working-instead-of-waiting style in an efficient event-driven model for I/O. In this tutorial we will describe Netty programming at high level and then transition to showing where other inefficiencies can still arise and how they can be addressed by pairing Netty with IO_uring and with hardware supported network steering capabilities.

Bios:

Dinesh Kumar is a Network Software Engineer in the Networking and Edge Group, working on Ethernet product development and programable Network solutions. as well om Network Technlogies enablement and optimizations for cloud and data center applications. He joined Intel in 1999 after completing his M.S in Computer Science from IET University, India.

Harshad Sane is a performance engineer in the Data Center and AI group with a deep technical expertise in system software, memory, and CPU architectures. He specializes in performance monitoring, software optimization, and tool development with focus in the cloud domain. Harshad joined Intel as a College Graduate in 2008 after completing his M.S. in Electrical and Computer Engineering from UC Boulder.

Rodolfo De Vega is a cloud software engineer in the Data Center and AI group, supporting customers and cloud service providers in optimizing their workloads for performance, both on-premises and in the public cloud. He joined Intel in 2000 after completing his B.S. in Computer Science from Saint Martin’s University.

Kshitij Doshi works at Intel Corporation in the Data Center and AI group, where he focuses on performance optimization of workloads and cloud instances. He obtained his undergraduate degree in Electrical Engineering from IIT Mumbai (1982) and his M.S. and Ph.D. degrees in Electrical and Computer Engineering from Rice University (1985 and 1989, respectively). His research interests span distributed systems, memory and storage architectures, and resource management.