Ordonnancement dans les systèmes d'exploitation

Dans les systèmes d'exploitation, l’ordonnanceur est le composant du noyau du système d'exploitation choisissant l'ordre d'exécution des processus sur les processeurs d'un ordinateur. En anglais, l'ordonnanceur est appelé scheduler. Un processus a besoin de la ressource processeur pour exécuter des calculs; il l'abandonne quand se produit une interruption, etc. De nombreux anciens processeurs ne peuvent effectuer qu'un traitement à la fois. Pour les autres, un ordonnanceur reste nécessaire pour déterminer quel processus sera exécuté sur quel processeur (c'est la notion d'affinité, très importante pour ne pas dégrader les performances). Au-delà des classiques processeurs multicœur, la notion d'hyperthreading rend la question de l'ordonnancement encore un peu plus complexe. À un instant donné, il y a souvent davantage de processus à exécuter que de processeurs. Un des rôles du système d'exploitation, et plus précisément de l'ordonnanceur du noyau, est de permettre à tous ces proces

Extending R2P2 with Congestion Control and Request-Level Scheduling

Konstantinos Prasopoulos

Distributed low-latency datacenter applications must meet strict service-level objectives in the order of microseconds. R2P2 is a novel datacenter transport protocol designed for such services. This work extends R2P2 with congestion control and request-level scheduling. It provides a configurable upper bound to in-network queuing using the idleness of the ToR-to-host link as the congestion signal. It also reduces the mean and tail RPC latency by minimizing head-of-line blocking in the network through shortest remaining processing time scheduling. The receiver driven scheduling mechanism is general and can support a variety of scheduling policies. The extension improves the performance of R2P2 under network load by at least 23% and up to two orders of magnitude. Finally, it enables R2P2 to perform similarly to a theoretically near-optimal, clean-slate transport protocol design.

2020

Energy-Efficient Co-Design Optimization of Many-Core Platforms for Big-Data Streaming Applications

Karim Kanoun

Big-Data streaming applications are used in several domains such as social media analysis, financial analysis, video annotation, surveillance, medical services and traffic prediction. These applications, running on different types of platforms from mobile devices to servers, are characterized by a highly-variable stochastic input data stream, stringent delay constraints and complex task graphs. Several software and hardware optimization techniques have been proposed to maximize the execution quality and the throughput of these applications and to minimize their energy consumption on many-core platforms. By analyzing the existing techniques, one can observe that most solutions classify as a hardware-based task-specific optimization, or as an operating system scheduler optimization, or yet as a load shedding mechanism at the application level. Each of these categories is limited in scope and can be blind to the nature of the program, the data being processed or the characteristics of the hardware. Big-Data streaming applications, due to their wide range of host hardware and content and dynamically-changing input streams, expose the fragmentation of the optimization techniques and create a clear need for a better approach. In this thesis, I propose a suite of energy-efficient hardware-software co-design techniques to bridge the gap between modern Big-Data streaming applications and existing many-core platforms. I choose to model the task graph of the class of applications I consider here by a direct acyclic graph (DAG). First, at the application layer, I propose a unified DAG monitoring solution to process online the general DAG model of the application and provide a set of relevant information that is leveraged at run time by a connected scheduler. At the operating system layer, I propose three different online scheduling solutions for many-core platforms which leverage the feedback from both the application and the hardware layers. The first scheduler addresses the problem of minimizing the energy consumption and the deadlines miss rates of multimedia applications. It takes advantage of the output of the DAG monitoring solution to adapt the mapping of the tasks of multimedia applications to the hardware according to the detected performance and targeted quality of service. The second and third schedulers address the problem of maximizing the quality and throughput and minimizing the energy consumption of Big-Data stream mining applications with single and multiple streams originating from different sources. To cope with the dynamically-changing Big-Data characteristics, the schedulers integrate machine-learning techniques to learn the environment dynamics and the application requirements and adapt the scheduling policy to the desired quality of service. I show that the proposed schedulers are able to scale the execution of data-mining applications to the system capability even in the presence of concept drift. Last, at the hardware layer, to address existing system architectures limitations and to increase even more the throughput, I propose a novel low-power many-core architecture for modern Big-Data stream mining applications that integrates a novel flexible memory hierarchy able to adapt to the dynamic data-driven nature of the input data stream.

EPFL2015

Energy-Efficient Co-Design Optimization of Many-Core Platforms for Big-Data Streaming Applications

Karim Kanoun

EPFL2015

Ordonnancement dans les systèmes d'exploitation

Extending R2P2 with Congestion Control and Request-Level Scheduling

A tool for an efficient comparison of scheduling methods

Energy-Efficient Co-Design Optimization of Many-Core Platforms for Big-Data Streaming Applications

Energy-Efficient Co-Design Optimization of Many-Core Platforms for Big-Data Streaming Applications

A tool for an efficient comparison of scheduling methods

Extending R2P2 with Congestion Control and Request-Level Scheduling