Computer Science

This paper examines disaggregated data center architectures from the perspective of the applications that would run on these data centers, and challenges the abstractions that have been proposed to date. In particular, we argue that operating systems for disaggregated data centers should not abstract disaggregated hardware resources, such as memory, compute, and storage away from applications, but should instead give them information about, and control over, these resources. To this end, we propose additional OS abstractions and interfaces for disaggregation and show how they can improve data transfer in data parallel frameworks and speed up failure recovery in replicated, fault-tolerant applications. This paper studies the technical challenges in providing applications with this additional functionality and advances several preliminary proposals to overcome these challenges.

Energy-efficient real-time task scheduling has been actively explored in the past decade. Different from the past work, this paper considers schedulability conditions for stochastic real-time tasks. A schedulability condition is first presented for frame-based stochastic real-time tasks, and several algorithms are also examined to check the schedulability of a given strategy. An approach is then proposed based on the schedulability condition to adapt a continuous-speed-based method to a discrete-speed system. The approach is able to stay as close as possible to the continuous-speed-based method, but still guaranteeing the schedulability. It is shown by simulations that the energy saving can be more than 20% for some system configurations

The objective of this paper is to take some aspects of disk scheduling and scheduling algorithms. The disk scheduling is discussed with a sneak peak in general and selection of algorithm in particular.

OS-level virtualization techniques virtualize system resources at the system call interface, has the distinct advantage of smaller run-time resource requirements as compared to HAL-level virtualization techniques, and thus forms an important building block for virtualizing parallel and distributed applications such as a HPC clusters. Because the Windows operating system puts certain critical functionalities in privileged user-level system service processes, a complete OS-level virtualization solution for the Windows platform requires duplication of such Windows service as Remote Procedure Call Server Service (RPCSS). As many implementation details of the Windows system services are proprietary, duplicating Windows system services becomes the key technical challenge for virtualizing the Windows platform at the OS level. Moreover, as a core component of cloud computing, IIS web server-related services need to be duplicated in containers (i.e., OS-level virtual machines), but so far there is no such scheme. In this paper, we thoroughly identify all issues that affect service duplication, and then propose the first known methodology to systematically duplicate both system and ordinary Windows services. Our experiments show that the methodology can duplicate a set of system and ordinary services on different versions of Windows OS.

With the availability of hybrid DRAM-NVRAM memory on the memory bus of CPUs, a number of file systems on NVRAM have been designed and implemented. In this paper we present the design and implementation of a file system on NVRAM called DurableFS, which provides atomicity and durability of file operations to applications. Due to the byte level random accessibility of memory, it is possible to provide these guarantees without much overhead. We use standard techniques like copy on write for data, and a redo log for metadata changes to build an efficient file system which provides durability and atomicity guarantees at the time a file is closed. Benchmarks on the implementation shows that there is only a 7 %degradation in performance due to providing these guarantees.

Energy consumption is a critical design issue in real-time systems, especially in battery- operated systems. Maintaining high performance, while extending the battery life between charges is an interesting challenge for system designers. Dynamic Voltage Scaling (DVS) allows a processor to dynamically change speed and voltage at run time, thereby saving energy by spreading run cycles into idle time. Knowing when to use full power and when not, requires the cooperation of the operating system scheduler. Usually, higher processor voltage and frequency leads to higher system throughput while energy reduction can be obtained using lower voltage and frequency. Instead of lowering processor voltage and frequency as much as possible, energy efficient real-time scheduling adjusts voltage and frequency according to some optimization criteria, such as low energy consumption or high throughput, while it meets the timing constraints of the real-time tasks. As the quantity and functional complexity of battery powered portable devices continues to raise, energy efficient design of such devices has become increasingly important. Many real-time scheduling algorithms have been developed recently to reduce energy consumption in the portable devices that use DVS capable processors. Three algorithms namely Red Tasks Only (RTO), Blue When Possible (BWP) and Red as Late as Possible (RLP) are proposed in the literature to schedule the real-time tasks in Weakly-hard real-time systems. This paper proposes optimal slack management algorithms to make the above existing weakly hard real-time scheduling algorithms energy efficient using DVS and DPD techniques.

We propose a framework that provides a programming interface to perform complex dynamic system-level analyses of deployed production systems. By leveraging hardware support for virtualization available nowadays on all commodity machines, our framework is completely transparent to the system under analysis and it guarantees isolation of the analysis tools running on its top. Thus, the internals of the kernel of the running system needs not to be modified and the whole platform runs unaware of the framework. Moreover, errors in the analysis tools do not affect the running system and the framework. This is accomplished by installing a minimalistic virtual machine monitor and migrating the system, as it runs, into a virtual machine. In order to demonstrate the potentials of our framework we developed an interactive kernel debugger, nicknamed HyperDbg. HyperDbg can be used to debug any critical kernel component, and even to single step the execution of exception and interrupt handlers.

Today's monolithic kernels often implement a small, fixed set of policies such as disk I/O scheduling policies, while exposing many parameters to let users select a policy or adjust the specific setting of the policy. Ideally, the parameters exposed should be flexible enough for users to tune for good performance, but in practice, users lack domain knowledge of the parameters and are often stuck with bad, default parameter settings. We present EOS, a system that bridges the knowledge gap between kernel developers and users by automatically evolving the policies and parameters in vivo on users' real, production workloads. It provides a simple policy specification API for kernel developers to programmatically describe how the policies and parameters should be tuned, a policy cache to make in-vivo tuning easy and fast by memorizing good parameter settings for past workloads, and a hierarchical search engine to effectively search the parameter space. Evaluation of EOS on four main Linux subsystems shows that it is easy to use and effectively improves each subsystem's performance.

Myopic is a hard real-time process scheduling algorithm that selects a suitable process based on a heuristic function from a subset (Window)of all ready processes instead of choosing from all available processes, like original heuristic scheduling algorithm. Performance of the algorithm significantly depends on the chosen heuristic function that assigns weight to different parameters like deadline, earliest starting time, processing time etc. and the sizeof the Window since it considers only k processes from n processes (where, k<= n). This research evaluates the performance of the Myopic algorithm for different parameters to demonstrate the merits and constraints of the algorithm. A comparative performance of the impact of window size in implementing the Myopic algorithm is presented and discussed through a set of experiments.

Data management applications store their data using structured files in which data are usually sorted to serve indexing and queries. In order to insert or remove a record in a sorted file, the positions of existing data need to be shifted. To this end, the existing data after the insertion or removal point must be rewritten to admit the change in place, which can be unaffordable for applications that make frequent updates. As a result, applications often employ extra layers of indirections to admit changes out-of-place. However, it causes increased access costs and excessive complexity. This paper presents a novel file abstraction, FlexFile, that provides a flexible file address space where in-place updates of arbitrary-sized data, such as insertions and removals, can be performed efficiently. With FlexFile, applications can manage their data in a linear file address space with minimal complexity. Extensive evaluation results show that a simple key-value store built on top of this abstraction can achieve high performance for both reads and writes.