Interoperable resource brokering with policy-based provisioning and job allocation

The increasing needs for computational power in areas such as weather simulation, genomics or Internet applications have led to sharing of geographically distributed and heterogeneous resources from commercial data centers and scientific institutions. Research in the areas of utility, grid and cloud computing, together with improvements in network and hardware virtualization has resulted in methods to locate and use resources to rapidly provision virtual environments in a flexible manner, while lowering costs for consumers and providers. ^ However, there is still a lack of methodologies to enable efficient and seamless sharing of resources among institutions. In this work, we concentrate in the problem of executing parallel scientific applications across distributed resources belonging to separate organizations. Our approach can be divided in three main points. First, we define and implement an interoperable grid protocol to distribute job workloads among partners with different middleware and execution resources. Second, we research and implement different policies for virtual resource provisioning and job-to-resource allocation, taking advantage of their cooperation to improve execution cost and performance. Third, we explore the consequences of on-demand provisioning and allocation in the problem of site-selection for the execution of parallel workloads, and propose new strategies to reduce job slowdown and overall cost.^

Approaches to Provisioning Network Topology of Virtual Machines in Cloud Systems

The current infrastructure as a service (IaaS) cloud systems, allow users to load their own virtual machines. However, most of these systems do not provide users with an automatic mechanism to load a network topology of virtual machines. In order to specify and implement the network topology, we use software switches and routers as network elements. Before running a group of virtual machines, the user needs to set up the system once to specify a network topology of virtual machines. Then, given the user’s request for running a specific topology, our system loads the appropriate virtual machines (VMs) and also runs separated VMs as software switches and routers. Furthermore, we have developed a manager that handles physical hardware failure situations. This system has been designed in order to allow users to use the system without knowing all the internal technical details.

Interoperable Resource Brokering with Policy-based Provisioning and Job Allocation

The increasing needs for computational power in areas such as weather simulation, genomics or Internet applications have led to sharing of geographically distributed and heterogeneous resources from commercial data centers and scientific institutions. Research in the areas of utility, grid and cloud computing, together with improvements in network and hardware virtualization has resulted in methods to locate and use resources to rapidly provision virtual environments in a flexible manner, while lowering costs for consumers and providers. However, there is still a lack of methodologies to enable efficient and seamless sharing of resources among institutions. In this work, we concentrate in the problem of executing parallel scientific applications across distributed resources belonging to separate organizations. Our approach can be divided in three main points. First, we define and implement an interoperable grid protocol to distribute job workloads among partners with different middleware and execution resources. Second, we research and implement different policies for virtual resource provisioning and job-to-resource allocation, taking advantage of their cooperation to improve execution cost and performance. Third, we explore the consequences of on-demand provisioning and allocation in the problem of site-selection for the execution of parallel workloads, and propose new strategies to reduce job slowdown and overall cost.

System design and resource management in the next-generation integrated WLAN and 3G cellular networks

Next-generation integrated wireless local area network (WLAN) and 3G cellular networks aim to take advantage of the roaming ability in a cellular network and the high data rate services of a WLAN. To ensure successful implementation of an integrated network, many issues must be carefully addressed, including network architecture design, resource management, quality-of-service (QoS), call admission control (CAC) and mobility management. ^ This dissertation focuses on QoS provisioning, CAC, and the network architecture design in the integration of WLANs and cellular networks. First, a new scheduling algorithm and a call admission control mechanism in IEEE 802.11 WLAN are presented to support multimedia services with QoS provisioning. The proposed scheduling algorithms make use of the idle system time to reduce the average packet loss of realtime (RT) services. The admission control mechanism provides long-term transmission quality for both RT and NRT services by ensuring the packet loss ratio for RT services and the throughput for non-real-time (NRT) services. ^ A joint CAC scheme is proposed to efficiently balance traffic load in the integrated environment. A channel searching and replacement algorithm (CSR) is developed to relieve traffic congestion in the cellular network by using idle channels in the WLAN. The CSR is optimized to minimize the system cost in terms of the blocking probability in the interworking environment. Specifically, it is proved that there exists an optimal admission probability for passive handoffs that minimizes the total system cost. Also, a method of searching the probability is designed based on linear-programming techniques. ^ Finally, a new integration architecture, Hybrid Coupling with Radio Access System (HCRAS), is proposed for lowering the average cost of intersystem communication (IC) and the vertical handoff latency. An analytical model is presented to evaluate the system performance of the HCRAS in terms of the intersystem communication cost function and the handoff cost function. Based on this model, an algorithm is designed to determine the optimal route for each intersystem communication. Additionally, a fast handoff algorithm is developed to reduce the vertical handoff latency.^

Fatty acids and stable isotopes as indicators of early-life feeding and potential maternal resource dependency in the bull shark Carcharhinus leucas

The degree of reliance of newborn sharks on energy reserves from maternal resource allocation and the timescales over which these animals develop foraging skills are critical factors towards understanding the ecological role of top predators in marine ecosystems. We used muscle tissue stable carbon isotopic composition and fatty acid analysis of bull sharks Carcharhinus leucas to investigate early-life feeding ecology in conjunction with maternal resource dependency. Values of δ13C of some young-of-the-year sharks were highly enriched, reflecting inputs from the marine-based diet and foraging locations of their mothers. This group of sharks also contained high levels of the 20:3ω9 fatty acid, which accumulates during periods of essential fatty acid deficiency, suggesting inadequate or undeveloped foraging skills and possible reliance on maternal provisioning. A loss of maternal signal in δ13C values occurred at a length of approximately 100 cm, with muscle tissue δ13C values reflecting a transition from more freshwater/estuarine-based diets to marine-based diets with increasing length. Similarly, fatty acids from sharks >100 cm indicated no signs of essential fatty acid deficiency, implying adequate foraging. By combining stable carbon isotopes and fatty acids, our results provided important constraints on the timing of the loss of maternal isotopic signal and the development of foraging skills in relation to shark size and imply that molecular markers such as fatty acids are useful for the determination of maternal resource dependency.

Scheduling medical application workloads on virtualized computing systems

This dissertation presents and evaluates a methodology for scheduling medical application workloads in virtualized computing environments. Such environments are being widely adopted by providers of "cloud computing" services. In the context of provisioning resources for medical applications, such environments allow users to deploy applications on distributed computing resources while keeping their data secure. Furthermore, higher level services that further abstract the infrastructure-related issues can be built on top of such infrastructures. For example, a medical imaging service can allow medical professionals to process their data in the cloud, easing them from the burden of having to deploy and manage these resources themselves. In this work, we focus on issues related to scheduling scientific workloads on virtualized environments. We build upon the knowledge base of traditional parallel job scheduling to address the specific case of medical applications while harnessing the benefits afforded by virtualization technology. To this end, we provide the following contributions: (1) An in-depth analysis of the execution characteristics of the target applications when run in virtualized environments. (2) A performance prediction methodology applicable to the target environment. (3) A scheduling algorithm that harnesses application knowledge and virtualization-related benefits to provide strong scheduling performance and quality of service guarantees. In the process of addressing these pertinent issues for our target user base (i.e. medical professionals and researchers), we provide insight that benefits a large community of scientific application users in industry and academia. Our execution time prediction and scheduling methodologies are implemented and evaluated on a real system running popular scientific applications. We find that we are able to predict the execution time of a number of these applications with an average error of 15%. Our scheduling methodology, which is tested with medical image processing workloads, is compared to that of two baseline scheduling solutions and we find that it outperforms them in terms of both the number of jobs processed and resource utilization by 20–30%, without violating any deadlines. We conclude that our solution is a viable approach to supporting the computational needs of medical users, even if the cloud computing paradigm is not widely adopted in its current form.

Improving caches in consolidated environments

Memory (cache, DRAM, and disk) is in charge of providing data and instructions to a computer's processor. In order to maximize performance, the speeds of the memory and the processor should be equal. However, using memory that always match the speed of the processor is prohibitively expensive. Computer hardware designers have managed to drastically lower the cost of the system with the use of memory caches by sacrificing some performance. A cache is a small piece of fast memory that stores popular data so it can be accessed faster. Modern computers have evolved into a hierarchy of caches, where a memory level is the cache for a larger and slower memory level immediately below it. Thus, by using caches, manufacturers are able to store terabytes of data at the cost of cheapest memory while achieving speeds close to the speed of the fastest one.^ The most important decision about managing a cache is what data to store in it. Failing to make good decisions can lead to performance overheads and over-provisioning. Surprisingly, caches choose data to store based on policies that have not changed in principle for decades. However, computing paradigms have changed radically leading to two noticeably different trends. First, caches are now consolidated across hundreds to even thousands of processes. And second, caching is being employed at new levels of the storage hierarchy due to the availability of high-performance flash-based persistent media. This brings four problems. First, as the workloads sharing a cache increase, it is more likely that they contain duplicated data. Second, consolidation creates contention for caches, and if not managed carefully, it translates to wasted space and sub-optimal performance. Third, as contented caches are shared by more workloads, administrators need to carefully estimate specific per-workload requirements across the entire memory hierarchy in order to meet per-workload performance goals. And finally, current cache write policies are unable to simultaneously provide performance and consistency guarantees for the new levels of the storage hierarchy.^ We addressed these problems by modeling their impact and by proposing solutions for each of them. First, we measured and modeled the amount of duplication at the buffer cache level and contention in real production systems. Second, we created a unified model of workload cache usage under contention to be used by administrators for provisioning, or by process schedulers to decide what processes to run together. Third, we proposed methods for removing cache duplication and to eliminate wasted space because of contention for space. And finally, we proposed a technique to improve the consistency guarantees of write-back caches while preserving their performance benefits.^

Mother–offspring isotope fractionation in two species of placentatrophic sharks

Stable-isotope values of a scalloped hammerhead Sphyrna lewini and blacktip shark Carcharhinus limbatus and their respective embryos were analysed. Embryos of both species were enriched in δ15N compared to their mothers (0·82 and 0·88‰, respectively), but fractionation of δ13C varied. Embryonic S. lewini were enriched (1·00‰) in δ13C while C. limbatus were depleted (0·27‰) relative to their mothers.

Host and network optimizations for performance enhancement and energy efficiency in data center networks

Modern data centers host hundreds of thousands of servers to achieve economies of scale. Such a huge number of servers create challenges for the data center network (DCN) to provide proportionally large bandwidth. In addition, the deployment of virtual machines (VMs) in data centers raises the requirements for efficient resource allocation and find-grained resource sharing. Further, the large number of servers and switches in the data center consume significant amounts of energy. Even though servers become more energy efficient with various energy saving techniques, DCN still accounts for 20% to 50% of the energy consumed by the entire data center. The objective of this dissertation is to enhance DCN performance as well as its energy efficiency by conducting optimizations on both host and network sides. First, as the DCN demands huge bisection bandwidth to interconnect all the servers, we propose a parallel packet switch (PPS) architecture that directly processes variable length packets without segmentation-and-reassembly (SAR). The proposed PPS achieves large bandwidth by combining switching capacities of multiple fabrics, and it further improves the switch throughput by avoiding padding bits in SAR. Second, since certain resource demands of the VM are bursty and demonstrate stochastic nature, to satisfy both deterministic and stochastic demands in VM placement, we propose the Max-Min Multidimensional Stochastic Bin Packing (M3SBP) algorithm. M3SBP calculates an equivalent deterministic value for the stochastic demands, and maximizes the minimum resource utilization ratio of each server. Third, to provide necessary traffic isolation for VMs that share the same physical network adapter, we propose the Flow-level Bandwidth Provisioning (FBP) algorithm. By reducing the flow scheduling problem to multiple stages of packet queuing problems, FBP guarantees the provisioned bandwidth and delay performance for each flow. Finally, while DCNs are typically provisioned with full bisection bandwidth, DCN traffic demonstrates fluctuating patterns, we propose a joint host-network optimization scheme to enhance the energy efficiency of DCNs during off-peak traffic hours. The proposed scheme utilizes a unified representation method that converts the VM placement problem to a routing problem and employs depth-first and best-fit search to find efficient paths for flows.

A Tale of Two Fishes: Using Recreational Angler Records to Examine the Link Between Fish Catches and Floodplain Connections in a Subtropical Coastal River

In the tropical and subtropical wet and dry regions, maintaining natural hydrologic connections between coastal rivers and adjacent ephemeral wetlands is critical to conserving and sustaining high levels of fisheries production within these systems. Though there is a consensus that there is a need to maintain these natural connections, little is known about what attributes of floodplain inundation regimes are most important in sustaining fisheries production. Two attributes of the flood season and thus floodplain inundation that may be particularly influential to fisheries are the amplitude of the flood season (floodplain water depth and spatial extent of inundation) and the duration of the flood season (i.e., time floodplains are inundated). In mangrove-dominated Everglades coastal rivers, seasonal inundation of upstream marsh floodplains may play an important role in provisioning recreational fisheries; however, this relationship remains unknown. Using two Everglades coastal river fisheries as a model, we tested whether the amplitude of the flood season or the duration of the flood season is more important in explaining variation in angler catch records of common snook and largemouth bass collected from 1992 to 2012. We validated angler catches with fisheries-independent electrofishing conducted in the same region from 2004 to 2012. Our results showed (1) that bass angler catches tracked electrofishing catches, while snook catches were completely mismatched. And (2) that previous year's marsh dynamics, particularly the duration of the flood season, was more influential than the flood season amplitude in explaining variation in bass catches, such that bass angler catches were negatively correlated to the period time that floodplains remained disconnected from coastal rivers in the previous year, while snook catches were not very well explained by floodplain inundation terms.

Simulation and analysis of network traffic for efficient and reliable information transfer

With the growing commercial importance of the Internet and the development of new real-time, connection-oriented services like IP-telephony and electronic commerce resilience is becoming a key issue in the design of TP-based networks. Two emerging technologies, which can accomplish the task of efficient information transfer, are Multiprotocol Label Switching (MPLS) and Differentiated Services. A main benefit of MPLS is the ability to introduce traffic-engineering concepts due to its connection-oriented characteristic. With MPLS it is possible to assign different paths for packets through the network. Differentiated services divides traffic into different classes and treat them differently, especially when there is a shortage of network resources. In this thesis, a framework was proposed to integrate the above two technologies and its performance in providing load balancing and improving QoS was evaluated. Simulation and analysis of this framework demonstrated that the combination of MPLS and Differentiated services is a powerful tool for QoS provisioning in IP networks.

Host and Network Optimizations for Performance Enhancement and Energy Efficiency in Data Center Networks

Modern data centers host hundreds of thousands of servers to achieve economies of scale. Such a huge number of servers create challenges for the data center network (DCN) to provide proportionally large bandwidth. In addition, the deployment of virtual machines (VMs) in data centers raises the requirements for efficient resource allocation and find-grained resource sharing. Further, the large number of servers and switches in the data center consume significant amounts of energy. Even though servers become more energy efficient with various energy saving techniques, DCN still accounts for 20% to 50% of the energy consumed by the entire data center. The objective of this dissertation is to enhance DCN performance as well as its energy efficiency by conducting optimizations on both host and network sides. First, as the DCN demands huge bisection bandwidth to interconnect all the servers, we propose a parallel packet switch (PPS) architecture that directly processes variable length packets without segmentation-and-reassembly (SAR). The proposed PPS achieves large bandwidth by combining switching capacities of multiple fabrics, and it further improves the switch throughput by avoiding padding bits in SAR. Second, since certain resource demands of the VM are bursty and demonstrate stochastic nature, to satisfy both deterministic and stochastic demands in VM placement, we propose the Max-Min Multidimensional Stochastic Bin Packing (M3SBP) algorithm. M3SBP calculates an equivalent deterministic value for the stochastic demands, and maximizes the minimum resource utilization ratio of each server. Third, to provide necessary traffic isolation for VMs that share the same physical network adapter, we propose the Flow-level Bandwidth Provisioning (FBP) algorithm. By reducing the flow scheduling problem to multiple stages of packet queuing problems, FBP guarantees the provisioned bandwidth and delay performance for each flow. Finally, while DCNs are typically provisioned with full bisection bandwidth, DCN traffic demonstrates fluctuating patterns, we propose a joint host-network optimization scheme to enhance the energy efficiency of DCNs during off-peak traffic hours. The proposed scheme utilizes a unified representation method that converts the VM placement problem to a routing problem and employs depth-first and best-fit search to find efficient paths for flows.