Capriccio

Abstract

This paper presents Capriccio, a scalable thread package for use with high-concurrency servers. While recent work has advocated event-based systems, we believe that threadbased systems can provide a simpler programming model that achieves equivalent or superior performance.By implementing Capriccio as a user-level thread package, we have decoupled the thread package implementation from the underlying operating system. As a result, we can take advantage of cooperative threading, new asynchronous I/O mechanisms, and compiler support. Using this approach, we are able to provide three key features: (1) scalability to 100,000 threads, (2) efficient stack management, and (3) resource-aware scheduling.We introduce linked stack management, which minimizes the amount of wasted stack space by providing safe, small, and non-contiguous stacks that can grow or shrink at run time. A compiler analysis makes our stack implementation efficient and sound. We also present resource-aware scheduling, which allows thread scheduling and admission control to adapt to the system's current resource usage. This technique uses a blocking graph that is automatically derived from the application to describe the flow of control between blocking points in a cooperative thread package. We have applied our techniques to the Apache 2.0.44 web server, demonstrating that we can achieve high performance and scalability despite using a simple threaded programming model.

The IX Operating System

“…The same design principles can benefit alternative, potentially application-specific, network protocols, as well as high-performance protocols for nonvolatile memory access. Finally, we will investigate library support for alternative APIs on top of our low-level interface, such as MegaPipe [Han et al 2012], cooperative threading [von Behren et al 2003], and rule-based models [Stutsman and Ousterhout 2013]. Such APIs and programming models will make it easier for applications to benefit from the performance and scalability advantages of IX.…”

DEICS: Data Erasure in Concurrent Software

“…Of the five, three of the applications (zebedee [5], retawq [41], mtdaapd [2]), handle sensitive data such as ftp passwords and database records. In order to further illustrate the precision and performance aspects of DEICS, we chose two additional applications (pfscan [3], knot [11]) from the RADAR benchmarks suite [4]. Table 3: Effect of transformation on Applications Table 2 shows the application sizes and transformation time taken.…”

Adaptive Task-Space Synchronization Control of SMMS Teleoperaiton Systems with Round-Robin Scheduling Protocol

“…There is a significant body of work on co-operative userspace thread libraries starting from Adya et al [20] who proposed the concept of userspace threading with automatic stack management. Capriccio [66] introduced optimizations to improve scalability and scheduling, and optimize stack allocation while identifying some of the pitfalls of non-preemptive scheduling. Inspired by scheduler activations [21], Arachne [58] made userspace threading coreaware and made thread creation very fast through careful engineering.…”