Typical schedulers in multi-tenancy environments make use of reactive, feedback-oriented mechanisms based on performance counters to avoid resource contention but suffer from detection lag and loss of performance. In this paper, we address these limitations by exploring the utility of predictive analysis through dynamic forecasting of applications' resource-heavy regions during its execution. Our compiler framework classifies loops in programs and leverages traditional compiler analysis along with learning mechanisms to quantify their behaviour. Based on the predictability of their execution time, it then inserts different types of beacons at their entry/exit points. The information produced by beacons in multiple processes is aggregated and analyzed by the proactive scheduler to respond to the anticipated workload requirements. For throughput environments, we develop a framework that demonstrates high-quality predictions and improvements in throughput over CFS by 76.78% on an average and up to 3.2x on Amazon Graviton2 Machine on consolidated workloads across 45 benchmarks.
Scheduling in Asymmetric Multicore Processors (AMP), a special case of Heterogeneous Multiprocessors, is a widely studied topic. The scheduling techniques which are mostly runtime do not usually consider parallel programming pattern used in parallel programming frameworks like OpenMP. On the other hand, current compilers for these parallel programming platforms are hardware oblivious which prevent any compile-time optimization for platforms like big.LITTLE and has to completely rely on runtime optimization. In this paper, we propose a hardware-aware Compiler Enhanced Scheduling (CES) where the…
We consider the problem of scheduling packets of different lengths via a directed communication link prone to jamming errors. Dynamic packet arrivals and errors are modelled by an adversary. We focus on estimating relative throughput of online scheduling algorithms, that is, the ratio between the throughputs achieved by the algorithm and the best scheduling for the same arrival and error patterns. This framework allows more accurate analysis of performance of online scheduling algorithms, even in worst-case arrival and error scenarios. We design an online algorithm for scheduling packets of ar…
Graphics processors, or GPUs, have recently been widely used as accelerators in the shared environments such as clusters and clouds. In such shared environments, many kernels are submitted to GPUs from different users, and throughput is an important metric for performance and total ownership cost. Despite the recently improved runtime support for concurrent GPU kernel executions, the GPU can be severely underutilized, resulting in suboptimal throughput. In this paper, we propose Kernelet, a runtime system with dynamic slicing and scheduling techniques to improve the throughput of concurrent ke…
Energy-aware architectures provide applications with a mix of low (LITTLE) and high (big) frequency cores. Choosing the best hardware configuration for a program running on such an architecture is difficult, because program parts benefit differently from the same hardware configuration. State-of-the-art techniques to solve this problem adapt the program's execution to dynamic characteristics of the runtime environment, such as energy consumption and throughput. We claim that these purely dynamic techniques can be improved if they are aware of the program's syntactic structure. To support this …
The unprecedented growth of data volumes has caused traditional approaches to computing to be re-evaluated. This has started a transition towards the use of very large-scale clusters of commodity hardware and has given rise to the development of many new languages and paradigms for data processing and analysis. In this paper, we propose a compiler technology-based alternative to the development of many different Big Data application infrastructures. Key to this approach is the development of a single intermediate representation that enables the integration of compiler optimization and query op…
We study throughput-optimum localized link scheduling in wireless networks. The majority of results on link scheduling assume binary interference models that simplify interference constraints in actual wireless communication. While the physical interference model reflects the physical reality more precisely, the problem becomes notoriously harder under the physical interference model. There have been just a few existing results on link scheduling under the physical interference model, and even fewer on more practical distributed or localized scheduling. In this paper, we tackle the challenges …
This paper investigates co-scheduling algorithms for processing a set of parallel applications. Instead of executing each application one by one, using a maximum degree of parallelism for each of them, we aim at scheduling several applications concurrently. We partition the original application set into a series of packs, which are executed one by one. A pack comprises several applications, each of them with an assigned number of processors, with the constraint that the total number of processors assigned within a pack does not exceed the maximum number of available processors. The objective i…
Data locality is a fundamental issue for data-parallel applications. Considering MapReduce in Hadoop, the map task scheduling part requires an efficient algorithm which takes data locality into consideration; otherwise, the system may become unstable under loads inside the system's capacity region and jobs may experience longer completion times which are not of interest. The data chunk needed for any map task can be in memory, on a local disk, in a local rack, in the same cluster or even in another data center. Hence, unless there has been much work on improving the speed of data center networ…
Modern computing platforms tend to deploy multiple GPUs (2, 4, or more) on a single node to boost system performance, with each GPU having a large capacity of global memory and streaming multiprocessors (SMs). GPUs are an expensive resource, and boosting utilization of GPUs without causing performance degradation of individual workloads is an important and challenging problem. Although services like MPS support simultaneous execution of multiple co-operative kernels on a single device, they do not solve the above problem for uncooperative kernels, MPS being oblivious to the resource needs of e…
We consider a natural scheduling problem which arises in many distributed computing frameworks. Jobs with diverse resource requirements (e.g. memory requirements) arrive over time and must be served by a cluster of servers, each with a finite resource capacity. To improve throughput and delay, the scheduler can pack as many jobs as possible in the servers subject to their capacity constraints. Motivated by the ever-increasing complexity of workloads in shared clusters, we consider a setting where the jobs' resource requirements belong to a very large number of diverse types or, in the extreme,…
We consider the classical problem of scheduling task graphs corresponding to complex applications on distributed computing systems. A number of heuristics have been previously proposed to optimize task scheduling with respect to metrics such as makespan and throughput. However, they tend to be slow to run, particularly for larger problem instances, limiting their applicability in more dynamic systems. Motivated by the goal of solving these problems more rapidly, we propose, for the first time, a graph convolutional network-based scheduler (GCNScheduler). By carefully integrating an inter-task …
Predicting the number of clock cycles a processor takes to execute a block of assembly instructions in steady state (the throughput) is important for both compiler designers and performance engineers. Building an analytical model to do so is especially complicated in modern x86-64 Complex Instruction Set Computer (CISC) machines with sophisticated processor microarchitectures in that it is tedious, error prone, and must be performed from scratch for each processor generation. In this paper we present Ithemal, the first tool which learns to predict the throughput of a set of instructions. Ithem…
OpenCL for FPGA enables developers to design FPGAs using a programming model similar for processors. Recent works have shown that code optimization at the OpenCL level is important to achieve high computational efficiency. However, existing works either focus primarily on optimizing single kernels or solely depend on channels to design multi-kernel pipelines. In this paper, we propose a source-to-source compiler framework, MKPipe, for optimizing multi-kernel workloads in OpenCL for FPGA. Besides channels, we propose new schemes to enable multi-kernel pipelines. Our optimizing compiler employs …
In this paper we explore the problem of achieving efficient packet transmission over unreliable links with worst case occurrence of errors. In such a setup, even an omniscient offline scheduling strategy cannot achieve stability of the packet queue, nor is it able to use up all the available bandwidth. Hence, an important first step is to identify an appropriate metric for measuring the efficiency of scheduling strategies in such a setting. To this end, we propose a relative throughput metric which corresponds to the long term competitive ratio of the algorithm with respect to the optimal. We …
The era of big data has led to the emergence of new systems for real-time distributed stream processing, e.g., Apache Storm is one of the most popular stream processing systems in industry today. However, Storm, like many other stream processing systems lacks an intelligent scheduling mechanism. The default round-robin scheduling currently deployed in Storm disregards resource demands and availability, and can therefore be inefficient at times. We present R-Storm (Resource-Aware Storm), a system that implements resource-aware scheduling within Storm. R-Storm is designed to increase overall thr…
