CPU Cycles Research Articles

Payload fragmentation framework for high-performance computing in cloud environment

Cloud computing environment delivers resources such as CPU cycles, storage, network service and memory through a web-based model. Users run their applications on remote infrastructure and get benefited more at a lesser cost. Compiling large-scale applications such as compiling GCC and compiling Clang in C and C++ environments where applications involve encryption breaking tools require extensive analysis. Lack of CPU cycles in consumer-grade computing will be detrimental to such applications. Fragmenting and distributing payloads across multiple clusters for such applications in a cloud-like environment is a challenging task. We have proposed a novel approach to perform payload distribution for users who want to run their computationally expensive tasks efficiently. The proposed framework has performed well compared to the ‘traditional’ payload execution policy.

Adaptive TB‐LMI: An efficient memory controller and scheduler design

SummaryIn the modern multi‐core systems, concurrently executing applications share common resource such as main memory. Memory scheduling algorithms are developed to resolve memory contention among competing applications so that throughput is high and fairness of the overall multi‐core system is guaranteed. In this paper, we present Adaptive Time‐based Least Memory Intensive (Adaptive TB‐LMI) scheduling, a new memory scheduling algorithm that addresses both fairness and system performance. Adaptive TB‐LMI is based on TB‐LMI which prioritizes applications according to their memory contention every pre‐defined CPU cycle. Adaptive TB‐LMI dynamically prioritizes applications according to their memory contention. Considering the previous algorithms with the best performance, for 16‐core system, TB‐LMI improves system throughput on average by 2.25X and 36% comparing to FCFS and TCM respectively. Adaptive TB‐LMI is 6% better than the TB‐LMI with static threshold. In terms of fairness and slowdown metrics, TB‐LMI show improvements of 30% and 3X, respectively, compared to FCFS and, 18% and 8%, respectively, compared to TCM. Adaptive TB‐LMI and TB‐LMI are 15% more efficient in energy‐delay product, although they are within 5% in terms of area overhead. Moreover, TCM has an area overhead of about 45% more than Adaptive TB‐LMI. In terms of Energy‐Delay Product, Adaptive TB‐LMI is 10% and 24% better than TB‐LMI and TCM, respectively. This is due to the dynamic capabilities of the adaptive algorithm to change the rate of the SQ, hence reducing the energy consumption.

The Definition and Software Performance of Hashstream, a Fast Length-Flexible PRF

Two of the fastest types of cryptographic algorithms are the stream cipher and the almost-universal hash function. There are secure examples of each that process data in software using less than one CPU cycle per byte. Hashstream combines the two types of algorithms in a straightforward manner yielding a PRF that can both consume inputs of and produce pseudorandom outputs of any desired length. The result is an object useful in many contexts: authentication, encryption, authenticated encryption, random generation, mask generation, etc. The HS1-SIV authenticated-encryption algorithm—a CAESAR competition second round selection—was based on Hashstream and showed the promise of such an approach by having provable security and topping the speed charts in several test configurations.

An Adaptive Approach Based on Resource-Awareness Towards Power-Efficient Real-Time Periodic Task Modeling on Embedded IoT Devices

Embedded devices are gaining popularity day by day due to the expanded use of Internet of Things applications. However, these embedded devices have limited capabilities concerning power and memory. Thus, the applications need to be tailored in such a way to perform the specified tasks within the constrained resources with the same accuracy. In Real-Time task scheduling, one of the challenging factors is the intelligent modelling of input tasks in such a way that it produces not only logically correct output within the deadline but also consumes minimum CPU power. Algorithms like Rate Monotonic and Earliest Deadline First compute hyper-period of input tasks for periodic repetition of the same set of tasks on CPU. However, at times when the tasks are not adequately modelled, they lead to an enormously high value of hyper-period which result in more CPU cycles and power consumption. Many state-of-the-art solutions are presented in this regard, but the main problem is that they limit tasks from having all possible period values; however, with the vision of Industry 4.0, where most of the tasks will be doing some critical manufacturing activities, it is highly discouraged to prevent them of a certain period. In this paper, we present a resource-aware approach to minimise the hyper-period of input tasks based on device profiles and allows tasks of every possible period value to admit. The proposed work is compared with similar existing techniques, and results indicate significant improvements regarding power consumptions.

Semi-Coherent DMA: An Alternative I/O Coherency Management for Embedded Systems

Many modern embedded CPUs adopt Non-Coherent DMA (NC-DMA) over Coherent DMA (C-DMA) because of simplicity. An NC-DMA design, however, requires a CPU device driver to explicitly invalidate or flush a wide range of cache space. When an I/O DMA device writes data to a main memory region, the CPU needs to invalidate the cache space corresponding to the same memory region twice: (1) to prevent dirty cache lines from overwriting the DMA data and (2) to remove any cache lines prefetched before the DMA is done. In this work, we first show that such explicit invalidations consume 31 percent of CPU cycles, limiting the data transfer throughput of a high-speed network interface card (NIC) when receiving network packets. Second, we propose a Semi-Coherent DMA (SC-DMA) architecture for improving the efficiency of NC-DMA with a slight modification to the hardware. Specifically, our SC-DMA records the DMA region and prohibits any data that is prefetched from the region from entering the cache, reducing nearly 50 percent of the unnecessary invalidations. Lastly, we identify several software optimizations that can substantially reduce excessive cache invalidations prevalent in NIC drivers. Our evaluation with NVIDIA Jetson TX2 shows that our proposed SC-DMA design with the NIC driver optimizations can improve the NIC data transfer throughput by up to 53.3 percent.

EPAS: A Sampling Based Similarity Identification Algorithm for the Cloud

The explosive growth of data brings new challenges to the data storage and management in cloud environment. These data usually have to be processed in a timely fashion in the cloud. Thus, any increased latency may cause a massive loss to the enterprises. Similarity detection plays a very important role in data management. Many typical algorithms such as Shingle, Simhash, Traits and Traditional Sampling Algorithm (TSA) are extensively used. The Shingle, Simhash and Traits algorithms read entire source file to calculate the corresponding similarity characteristic value, thus requiring lots of CPU cycles and memory space and incurring tremendous disk accesses. In addition, the overhead increases with the growth of data set volume and results in a long delay. Instead of reading entire file, TSA samples some data blocks to calculate the fingerprints as similarity characteristics value. The overhead of TSA is fixed and negligible. However, a slight modification of source files will trigger the bit positions of file content shifting. Therefore, a failure of similarity identification is inevitable due to the slight modifications. This paper proposes an Enhanced Position-Aware Sampling algorithm (EPAS) to identify file similarity for the cloud by modulo file length. EPAS concurrently samples data blocks from the head and the tail of the modulated file to avoid the position shift incurred by the modifications. Meanwhile, an improved metric is proposed to measure the similarity between different files and make the possible detection probability close to the actual probability. Furthermore, this paper describes a query algorithm to reduce the time overhead of similarity detection. Our experimental results demonstrate that the EPAS significantly outperforms the existing well known algorithms in terms of time overhead, CPU and memory occupation. Moreover, EPAS makes a more preferable tradeoff between precision and recall than that of other similarity detection algorithms. Therefore, it is an effective approach of similarity identification for the cloud.

HULK and DDoS Attacks in Web Applications with Detection Mechanism

Distributed Denial of Service attacks are significant dangers these days over web applications and web administrations. These assaults pushing ahead towards application layer to procure furthermore, squander most extreme CPU cycles. By asking for assets from web benefits in gigantic sum utilizing quick fire of solicitations, assailant robotized programs use all the capacity of handling of single server application or circulated environment application. The periods of the plan execution is client conduct checking and identification. In to beginning with stage by social affair the data of client conduct and computing individual user’s trust score will happen and Entropy of a similar client will be ascertained. HTTP Unbearable Load King (HULK) attacks are also evaluated. In light of first stage, in recognition stage, variety in entropy will be watched and malevolent clients will be recognized. Rate limiter is additionally acquainted with stop or downsize serving the noxious clients. This paper introduces the FAÇADE layer for discovery also, hindering the unapproved client from assaulting the framework.

An XML based Web Crawler with Page Revisit Policy and Updation in Local Repository of Search Engine

In a large collection of web pages, it is difficult for search engines to keep their online repository updated. Major search engines have hundreds of web crawlers that crawl the WWW day and night and send the downloaded web pages via a network to be stored in the search engine’s database. These results in over utilization of network resources like bandwidth, CPU cycles and so on. This paper proposes an architecture that tries to reduce the utilization of shared network resources with the help of an advanced XML based approach. This focused crawling based architecture is trained to download only the high quality data from the internet leaving behind the web pages which are not relevant to the desired domain. Here, a detailed layout of the proposed system is described which is capable of reducing the load on network and reducing the problem arise in residency of mobile agent at the remote server.

A server-based approach for predictable GPU access with improved analysis

Abstract We propose a server-based approach to manage a general-purpose graphics processing unit (GPU) in a predictable and efficient manner. Our proposed approach introduces a GPU server that is a dedicated task to handle GPU requests from other tasks on their behalf. The GPU server ensures bounded time to access the GPU, and allows other tasks to suspend during their GPU computation to save CPU cycles. By doing so, we address the two major limitations of the existing real-time synchronization-based GPU management approach: busy waiting within critical sections and long priority inversion. We have implemented a prototype of the server-based approach on a real embedded platform. This case study demonstrates the practicality and effectiveness of the server-based approach. Experimental results indicate that the server-based approach yields significant improvements in task schedulability over the existing synchronization-based approach in most practical settings. Although we focus on a GPU in this paper, the server-based approach can also be used for other types of computational accelerators.

Multi-prediction based scheduling for hybrid workloads in the cloud data center

Cloud computing can leverage over-provisioned resources that are wasted in traditional data centers hosting production applications by consolidating tasks with lower QoS and SLA requirements. However, the dramatic fluctuation of workloads with lower QoS and SLA requirements may impact the performance of production applications. Frequent task eviction, killing and rescheduling operations also waste CPU cycles and create overhead. This paper aims to schedule hybrid workloads in the cloud data center to reduce task failures and increase resource utilization. The multi-prediction model, including the ARMA model and the feedback based online AR model, is used to predict the current and the future resource availability. Decision to accept or reject a new task is based on the available resources and task properties. Evaluations show that the scheduler can reduce the host overload and failed tasks by nearly 70%, and increase effective resource utilization by more than 65%. The task delay performance degradation is also acceptable.

Optimal Network Service Chain Provisioning

Service chains consist of a set of network services, such as firewalls or application delivery controllers, which are interconnected through a network to support various applications. While it is not a new concept, there has been an extremely important new trend with the rise of software-defined network (SDN) and Network Function Virtualization (NFV). The combination of SDN and NFV can make the service chain and application provisioning process much shorter and simpler. In this paper, we study the provisioning of service chains jointly with the number/location of virtual network functions (VNFs). While chains are often built to support multiple applications, the question arises as how to plan the provisioning of service chains in order to avoid data passing through unnecessary network devices or servers and consuming extra bandwidth and CPU cycles. It requires choosing carefully the number and the location of the VNFs. We propose an exact mathematical model using decomposition methods whose solution is scalable in order to conduct such an investigation. We conduct extensive numerical experiments, and show we can solve exactly the routing of service chain requests in a few minutes for networks with up to 50 nodes, and traffic requests between all pairs of nodes. Detailed analysis is then made on the best compromise between minimizing the bandwidth requirement and minimizing the number of VNFs and optimizing their locations using different data sets.

Resource Provisioning of UID Application in Cloud Computing Domain

Resource provisioning is the sharing and distribution of resources among various applications deployed on a server. The resource provisioning is usually performed either at server level or at client level. Server uses resource provisioning to determine which application is going to receive a resource when many applications request for the same resource at the same time. At client level, resource provisioning is the process to determine which client will be granted an access to the application, when the number of clients requesting for access to an application is more than the available instances of the application. The resources an application requests for includes, buffer space, time to complete its execution, battery life in wireless devices, CPU cycles et al. In this work, cloud computing was used since it provides a facility of large resource capacity which can be shared and distributed amongst millions of requesting client applications. Therefore, it is possible to execute more than one application at the same time. The applications are granted resources based on certain scheduling algorithms. A cloud user can access the cloud services on pay-per-use basis, in case of dynamic provisioning. In dynamic provisioning the requirements of an application may vary during the course of its execution. In this work, dynamic provisioning was implemented. To show the effectiveness of Resource Provisioning a UID (Unique Identifier), an Open Source Application has been successfully designed and deployed on open source Cloud Server.

A fabric architecture towards block chain application using hyper ledger

Block chain is a trust which can be best understood by the state machine replication, It is decentralized distributed ledger which is along all peers in the network connected through nodes over Internet. Every Node in the chain have equal stake and also the main factor is every node which have greater CPU cycles has a chance to operate node by spending those CPU cycles and also show Proof Of Work. Hence when the block chain is used in Business Models it loses its Private transactions and Confidential Contracts.

Resource provisioning methodology for cloud environment with producer and consumer favorable: an approach of virtual cloud compiler

Cloud computing is a model where traditional resources such as CPU cycles, storage, security etc. are delivered through web based. It is a technology which has ability to change large part of software development cycle, 3D rendering or any other computationally expensive tasks execution. Much amount of time is wasted on compiling and rendering such computationally expensive tasks due to low power machines, which directly proportional to efficiency of user who is working on that project. Extreme computational tasks such as weather forecast, DNA analyses, encryption breaking takes so much time in consumer grade computing devices that they are realistically not possible to perform. We have proposed a novel approach to perform payload distribution, for the users who wanted to run their computationally expensive tasks efficiently. We have used virtualization technique on data center resources to perform scheduling. Up to 32% cost has been reduced in an environment of 30 users when our technology used instead of traditional standalone desktop environment. This is achieved by replacing 30 standalone computers with a powerful server and thin clients like Raspberry pi as clients. Time wasted in computational task such as rendering and compiling is greatly reduced. We have not only improved the efficiency, but also make sure both cloud producer and consumer are favorable. With simulations and outcomes, we validate that our methodology for payload distribution performs well.

A Study of Relations between Optimal Network Card Mode and Data Transmission Size

There is a necessity to study the relations between optimal network card mode and its data transmission size, in order to investigate if a specific network card mode is more suitable for a certain size of transmission. Three network card modes are examined, which are active wait, passive wait, and watermark. The existing data of 47 different transmission sizes and their respective optimal network card modes are observed. The analyses conclude that all network card modes are suitable for all levels of transmission size (small, medium, and big), as long as their optimal value takes place. However, it is better to implement the network card mode with the least generated kernel interrupts. Hence, the CPU cycles can be preserved.

The performance cost of software obfuscation for Android applications

Software security of Android applications is especially susceptible (vulnerable) to malicious reverse engineer than the native code based software, because its Java bytecode is easier to decompile and to reconstruct the original Java source code. Therefore obfuscation is an essential criterion for the protection of Android applications. Meanwhile, the unpredicted performance loss will be caused by the obfuscation, which might seriously influence the user experience of the software. Therefore the obfuscation degree and the performance cost out of it require being optimized. In this paper, we are working on the problem that: to obfuscate an Android application to a target level of “difficulty” degree, while not substantially “slowing it down”. We measure “difficulty” by utilizing software complexity metrics and the “slow” in CPU cycles. Within the framework designed for obfuscating Android applications, we implement the “Naïve Bayesian Classifier” algorithm for the optimized obfuscation of the software complexity and the performance, and show that it outperforms the algorithm whose predictions rely on mean values. We furthermore investigate the performance penalty imposed by obfuscation process when framework targets different complexity values and metrics. We show that some obfuscation methods are more performance costly than others to achieve the same metric value increase. Our result shows that, for any given software complexity, the required performance benefits can be achieved if the right obfuscation techniques are used.

ATLAS Distributed Computing Experience and Performance During the LHC Run-2

ATLAS Distributed Computing during LHC Run-1 was challenged by steadily increasing computing, storage and network requirements. In addition, the complexity of processing task workflows and their associated data management requirements led to a new paradigm in the ATLAS computing model for Run-2, accompanied by extensive evolution and redesign of the workflow and data management systems. The new systems were put into production at the end of 2014, and gained robustness and maturity during 2015 data taking. ProdSys2, the new request and task interface; JEDI, the dynamic job execution engine developed as an extension to PanDA; and Rucio, the new data management system, form the core of Run-2 ATLAS distributed computing engine.One of the big changes for Run-2 was the adoption of the Derivation Framework, which moves the chaotic CPU and data intensive part of the user analysis into the centrally organized train production, delivering derived AOD datasets to user groups for final analysis. The effectiveness of the new model was demonstrated through the delivery of analysis datasets to users just one week after data taking, by completing the calibration loop, Tier-0 processing and train production steps promptly. The great flexibility of the new system also makes it possible to execute part of the Tier-0 processing on the grid when Tier-0 resources experience a backlog during high data-taking periods.The introduction of the data lifetime model, where each dataset is assigned a finite lifetime (with extensions possible for frequently accessed data), was made possible by Rucio. Thanks to this the storage crises experienced in Run-1 have not reappeared during Run-2. In addition, the distinction between Tier-1 and Tier-2 disk storage, now largely artificial given the quality of Tier-2 resources and their networking, has been removed through the introduction of dynamic ATLAS clouds that group the storage endpoint nucleus and its close-by execution satellite sites. All stable ATLAS sites are now able to store unique or primary copies of the datasets.ATLAS Distributed Computing is further evolving to speed up request processing by introducing network awareness, using machine learning and optimisation of the latencies during the execution of the full chain of tasks. The Event Service, a new workflow and job execution engine, is designed around check-pointing at the level of event processing to use opportunistic resources more efficiently.ATLAS has been extensively exploring possibilities of using computing resources extending beyond conventional grid sites in the WLCG fabric to deliver as many computing cycles as possible and thereby enhance the significance of the Monte-Carlo samples to deliver better physics results.The exploitation of opportunistic resources was at an early stage throughout 2015, at the level of 10% of the total ATLAS computing power, but in the next few years it is expected to deliver much more. In addition, demonstrating the ability to use an opportunistic resource can lead to securing ATLAS allocations on the facility, hence the importance of this work goes beyond merely the initial CPU cycles gained.In this paper, we give an overview and compare the performance, development effort, flexibility and robustness of the various approaches.

Using OSG Computing Resources with (iLC)Dirac

CPU cycles for small experiments and projects can be scarce, thus making use of all available resources, whether dedicated or opportunistic, is mandatory. While enabling uniform access to the LCG computing elements (ARC, CREAM), the DIRAC grid interware was not able to use OSG computing elements (GlobusCE, HTCondor-CE) without dedicated support at the grid site through so called ‘SiteDirectors’, which directly submit to the local batch system. This in turn requires additional dedicated effort for small experiments on the grid site. Adding interfaces to the OSG CEs through the respective grid middleware is therefore allowing accessing them within the DIRAC software without additional site-specific infrastructure. This enables greater use of opportunistic resources for experiments and projects without dedicated clusters or an established computing infrastructure with the DIRAC software. To allow sending jobs to HTCondor-CE and legacy Globus computing elements inside DIRAC the required wrapper classes were developed. Not only is the usage of these types of computing elements now completely transparent for all DIRAC instances, which makes DIRAC a flexible solution for OSG based virtual organisations, but it also allows LCG Grid Sites to move to the HTCondor-CE software, without shutting DIRAC based VOs out of their site. In these proceedings we detail how we interfaced the DIRAC system to the HTCondor-CE and Globus computing elements and explain the encountered obstacles and solutions developed, and how the linear collider community uses resources in the OSG.

Formal Verification of a Timing Enforcer Implementation

A timing enforcer is a scheduler that not only allocates CPU cycles to threads, but also uses timers to enforce time budgets. An approach for verifying safety properties of timing enforcers at the source code level is presented. We assume that the enforcer is implemented as a set of “enforcer” functions that are executed atomically on critical system-level events, such as the arrival and departure of jobs, and triggering of timers. The key idea is to express the safety property as an invariant, and prove that it is inductive across all the enforcer functions. A formal semantics of timing enforcers is presented, including the semantics of functions used to read the system clock and set timers. Using this semantics, the verification approach is presented, and its soundness proved. Further, the approach also takes into consideration the periodicity of tasks. It is validated by proving the correctness of the enforcement of CPU cycle budgets for tasks by the Zero-Slack Rate Monotonic ( zsrm ) scheduler, which is implemented in C as a Linux kernel module. The inductiveness of the necessary zsrm invariants is proved by expressing them as function contracts using the acsl specification language, and verifying the contracts using the frama-c tool.

Contract-Based Verification of Timing Enforcers

A timing enforcer not only allocates CPU cycles to threads but also uses timers to enforce time budgets. An approach for verifying safety properties of timing enforcers at the source code level is presented. We assume that the enforcer is implemented as a set of entry functions that are executed atomically on critical system-level events, such as arrival and departure of periodic jobs. The key idea is to express the safety property as an invariant, and prove that it is inductive across all the entry functions. The approach is validated by proving correctness of the enforcement of CPU cycle budgets for tasks by a mixed-criticality scheduler called zsrm that is implemented in C. The inductiveness of the necessary zsrm invariants is proved by expressing them as function contracts using the acsl specification language, and verifying the contracts using the frama-c tool.