1,721,053 research outputs found
BOSP: BarbequeRTRM Open Source Project
No full textThe BarbequeRTRM is a framework being developed at DEI - Politecnico di Milano - under the context of the European Project 2PARMA and it has been partially funded by the EC under the FP7-ICT-248716-2PARMA grant. This framework is the core of an highly modular and extensible run-time resource manager which provide support for an easy integration and management of multiple applications competing on the usage of one (or more) shared MIMD many-core computation devices. The framework design, which exposes different plugin interfaces, provides support for pluggable policies for both resource scheduling and the management of applications coordination and reconfiguration.
Applications integrated with this framework gets “for-free” a suitable instrumentation to support Design-Space-Exploration (DSE) techniques, which could be used to profile application behaviors to either optimize them at design time or support the identification of optimal QoS requirements goals as well as their run-time monitoring.
Suitable platform abstraction layers, built on top of Linux kernel interfaces, allows an easily porting of the framework on different platforms and its integration with specific execution environments such as the Android run-time.
Based on all these features the framework allows an easily coding of resource management policies which support an optimized assignment of resources to demanding applications considering:
- application properties, e.g. run-time requirements, operating modes and relative priorities,
- resources availability and state, e.g. power and thermal conditions
- tunable run-time optimization goals, e.g. power reduction, energy optimization, reconfiguration overheads minimization and overall performances maximization.
An initial version of the proposed framework is already available and actively developed as an OpenSource project, the Barbeque OpenSource Project (BOSP), which defines a new approach to develop a System-Wide RTRM supporting a comprehensive set of advanced features, such as:
- a hierarchical and distributed control
- the exploitation of design-time information
- a rich multi-objective optimization strategy
- a portable and modular design based on a set of tunable policies
Co-scheduling tasks on multi-core heterogeneous systems: An energy-aware perspective
No full textSingle-ISA heterogeneous multi-core processors trade-off power with performance; however, threads that co-
run on shared resources suffer from resource contention, which induces performance degradation and energy
inefficiency. The authors introduce a novel approach to optimise the co-scheduling of multi-threaded applications on
heterogeneous processors. The approach is based on the concept of stakes function, which represents the trade-off
between isolation and sharing of resources. The authors also develop a co-scheduling algorithm that use stakes
functions to optimise resource usage while mitigating resource contention, thus improving performance and energy
efficiency. They validated the approach using applications from the Princeton Application Repository for Shared-
Memory Computers (PARSEC) benchmark suite, obtaining up to 12.88% performance speed-up, 13.65% energy speed-
up and 28.29% energy delay speed-up with respect to the standard Linux heterogeneous multi-processing scheduler
The Misconception of Exponential Tail Upper-Bounding in Probabilistic Real-Time
Full text linkMeasurement-Based Probabilistic Timing Analysis, a probabilistic real-time computing method, is based on the Extreme Value Theory (EVT), a statistical theory applied to Worst-Case Execution Time analysis on real-time embedded systems. The output of the EVT theory is a statistical distribution, in the form of Generalized Extreme Value Distribution or Generalized Pareto Distribution. Their cumulative distribution function can asymptotically assume one of three possible forms: light, exponential or heavy tail. Recently, several works proposed to upper-bound the light-tail distributions with their exponential version. In this paper, we show that this assumption is valid only under certain conditions and that it is often misinterpreted. This leads to unsafe estimations of the worst-case execution time, which cannot be accepted in applications targeting safety critical embedded systems
Timing Predictability in High-Performance Computing with Probabilistic Real-Time
Full text linkApplication requirements in High-Performance Computing (HPC) are becoming increasingly exacting, and the demand for computational resources is rising. In parallel, new application domains are emerging, as well as additional requirements, such as meeting real-time constraints. This requirement, typical of embedded systems, is difficult to guarantee when dealing with HPC infrastructures, due to the intrinsic complexity of the system. Traditional embedded systems static analyses to estimate the Worst-Case Execution Time (WCET) are not applicable to HPC, because modeling and analyzing all the system’s hardware and software components is not practical. Measurement-based probabilistic analyses for the WCET emerged in the last decade to overcome these issues, but it requires the system to satisfy certain conditions to estimate a correct and safe WCET. In this work, we show the emerging application timing requirements, and we propose to exploit the probabilistic real-time theory to achieve the required time predictability. After a brief recap of the fundamentals of this methodology, we focus on its applicability to HPC systems, to check their ability to satisfy such conditions. In particular, we studied the advantages of having heterogeneous processors in HPC nodes and how resource management affects the applicability of the proposed technique
Effective runtime resource management using linux control groups with the BarbequeRTRM framework
Full text linkThe extremely high technology process reached by silicon manufacturing (smaller than 32nm) has led to production of computational platforms and SoC, featuring a considerable amount of resources. Whereas from one side such multi- and many-core platforms show growing performance capabilities, from the other side they are more and more affected by power, thermal, and reliability issues. Moreover, the increased computational capabilities allows congested usage scenarios with workloads subject to mixed and time-varying requirements. Effective usage of the resources should take into account both the application requirements and resources availability, with an arbiter, namely a resource manager in charge to solve the resource contention among demanding applications.
Current operating systems (OS) have only a limited knowledge about application-specific behaviors and their time-varying requirements. Dedicated system interfaces to collect such inputs and forward them to the OS (e.g., its scheduler) are thus an interesting research area that aims at integrating the OS with an ad hoc resource manager. Such a component can exploit efficient low-level OS interfaces and mechanisms to extend its capabilities of controlling tasks and system resources. Because of the specific tasks and timings of a resource manager, this component can be easily and effectively developed as a user-space extension lying in between the OS and the controlled application.
This article, which focuses on multicore Linux systems, shows a portable solution to enforce runtime resource management decisions based on the standard control groups framework. A burst and a mixed workload analysis, performed on a multicore-based NUMA platform, have reported some promising results both in terms of performance and power saving
Probabilistic-WCET Reliability: Statistical Testing of EVT hypotheses
Full text linkIn recent years, the interest in probabilistic real-time has grown, as a response to the limitations of traditional static Worst-Case Execution Time (WCET) methods, in performing timing analysis of applications running on complex systems, like multi/many-cores and COTS platforms. The probabilistic theory can partially solve this problem, but it requires strong guarantees on the execution time traces, in order to provide safe probabilistic-WCET estimations. These requirements can be verified through suitable statistical tests, as described in this paper. In this work, we identify also challenges and problems of using statistical testing procedures in probabilistic real-time computing, proposing a unified test procedure based on a single index called Probabilistic Predictability Index (PPI). An experimental campaign has been carried out, considering both synthetic and realistic datasets, and the analysis of the impact of the Linux PREEMPT_RT patch on a modern complex platform as a use-case of the proposed index
Resource-Aware Application Execution Exploiting the BarbequeRTRM
Full text linkEnergy efficiency and thermal management have become ma- jor concerns in both embedded and HPC systems. The progress of silicon technology and the subsequent growth of the dark silicon phenomena are negatively a ecting the reliability of computing systems. As a result, in the next future we expect run-time variability to increase in terms of both performance and computing resources availability. To address these is- sues, systems and applications must be able to adapt to such scenarios. This work provides a brief overview of the Barbeque Run-Time Resource Manager (BarbequeRTRM) and the application execution model that it exploits, in order to deal with run-time performance and available re- sources variability
- …
