Umeå University's logo

We present the Disaster Recovery Layer (DRL) that enables OpenStack-managed datacenter workloads, Virtual Machines (VMs) and Volumes, to be protected and recovered in another datacenter, in case of a disaster. This work has been carried out in the context of the EU FP7 ORBIT project that develops technologies for enabling business continuity as a service. The DRL framework is based on a number of autonomous components and extensions of OpenStack modules, while its functionalities are available through OpenStack's Horizon UI and command line interface. Also, the DRL's architecture is extensible, allowing for the easy and dynamic integration of protection, restoration and orchestration plug-ins that adopt new approaches. A distributed disaster detection mechanism was also developed for identifying datacenter disasters and alerting the DRL. For the evaluation of the DRL, a two (active and backup) datacenters testbed has been setup in respective sites in Umea and Lulea, 265km apart and connected through the Swedish national research and education network. In case of a disaster, traffic is redirected between the datacenters utilizing the BGP anycast scheme. The experiments performed, show that DRL can efficiently protect VMs and Volumes, with minimum service disruption in case of failures and low overhead, even when the available bandwidth is limited.

Active Virtual Machine (VM) replication is an application independent and cost-efficient mechanism for high availability and fault tolerance, with several recently proposed implementations based on checkpointing. However, these methods may suffer from large impacts on application latency, excessive resource usage overheads, and/or unpredictable behavior for varying workloads. To address these problems, we propose a hybrid approach through a Proportional-Integral (PI) controller to dynamically switch between periodic and on-demand check-pointing. Our mechanism automatically selects the method that minimizes application downtime by adapting itself to changes in workload characteristics. The implementation is based on modifications to QEMU, LibVirt, and OpenStack, to seamlessly provide fault tolerant VM provisioning and to enable the controller to dynamically select the best checkpointing mode. Our evaluation is based on experiments with a video streaming application, an e-commerce benchmark, and a software development tool. The experiments demonstrate that our adaptive hybrid approach improves both application availability and resource usage compared to static selection of a checkpointing method, with application performance gains and neglectable overheads.

Energy management has become increasingly necessary in data centers to address all energy-related costs, including capital costs, operating expenses, and environmental impacts. Heterogeneous systems with mixed hardware architectures provide both throughput and processing efficiency for different specialized application types and thus have a potential for significant energy savings. However, the presence of multiple and different processing elements increases the complexity of resource assignment. In this paper, we propose a system for efficient resource management in heterogeneous clouds. The proposed approach maps applications' requirement to different resources reducing power usage with minimum impact on performance. A technique that combines the scheduling of custom hardware accelerators, in our case, Field-Programmable Gate Arrays (FPGAs) and optimized resource allocation technique for commodity servers, is proposed. We consider an energy-aware scheduling technique that uses both the applications' performance and their deadlines to control the assignment of FPGAs to applications that would consume the most energy. Once the scheduler has performed the mapping between a VM and an FPGA, an optimizer handles the remaining VMs in the server, using vertical scaling and CPU frequency adaptation to reduce energy consumption while maintaining the required performance. Our evaluation using interactive and data-intensive applications compare the effectiveness of the proposed solution in energy savings as well as maintaining applications performance, obtaining up to a 32% improvement in the performance-energy ratio on a mix of multimedia and e-commerce applications.

Despite their proliferation as a dominant computing paradigm, cloud computing systems lack effective mechanisms to manage their vast amounts of resources efficiently. Resources are stranded and fragmented, ultimately limiting cloud systems' applicability to large classes of critical applications that pose non-moderate resource demands. Eliminating current technological barriers of actual fluidity and scalability of cloud resources is essential to strengthen cloud computing's role as a critical cornerstone for the digital economy. ACTiCLOUD proposes a novel cloud architecture that breaks the existing scale-up and share-nothing barriers and enables the holistic management of physical resources both at the local cloud site and at distributed levels. Specifically, it makes advancements in the cloud resource management stacks by extending state-of-the-art hypervisor technology beyond the physical server boundary and localized cloud management system to provide a holistic resource management within a rack, within a site, and across distributed cloud sites. On top of this, ACTiCLOUD will adapt and optimize system libraries and runtimes (e.g., JVM) as well as ACTiCLOUD-native applications, which are extremely demanding, and critical classes of applications that currently face severe difficulties in matching their resource requirements to state-of-the-art cloud offerings.

Resource sharing is an inherent characteristic of cloud data centers. Virtual Machines (VMs) and/or Containers that are co-located in the same physical server often compete for resources leading to interference. The noisy neighbor’s effect refers to an anomaly caused by a VM/container limiting resources accessed by another one. Our main contribution is an online, lightweight and application-agnostic solution for anomaly detection, that follows an unsupervised approach. It is based on comparing models for different lags: Dirichlet Process Gaussian Mixture Models to characterize the resource usage profile of the application, and distance measures to score the similarity among models. An alarm is raised when there is an abrupt change in short-term lag (i.e. high distance score for short-term models), while the long-term state remains constant. We test the algorithm for different cloud workloads: websites, periodic batch applications, Spark-based applications, and Memcached server. We are able to detect anomalies in the CPU and memory resource usage with up to 82–96% accuracy (recall) depending on the scenario. Compared to other baseline methods, our approach is able to detect anomalies successfully, while raising low number of false positives, even in the case of applications with unusual normal behavior (e.g. periodic). Experiments show that our proposed algorithm is a lightweight and effective solution to detect noisy neighbor effect without any historical info about the application, that could also be potentially applied to other kind of anomalies.

A lack of energy proportionality, low resource utilization, and interference in virtualized infrastructure make the cloud a challenging target environment for improving energy efficiency. In this paper we present OptiBook, a system that improves energy proportionality and/or resource utilization to optimize performance and energy efficiency. OptiBook shares servers between latency-sensitive services and batch jobs, over- books the system in a controllable manner, uses vertical (CPU and DVFS) scaling for prioritized virtual machines, and applies performance isolation techniques such as CPU pinning and quota enforcement as well as online resource tuning to effectively improve energy efficiency. Our evaluations show that on average, OptiBook improves performance per watt by 20% and reduces energy consumption by 9% while minimizing SLO violations. 

Complete data center failures may occur due to disastrous events such as earthquakes or fires. To attain robustness against such failures and reduce the probability of data loss, data must be replicated in another data center sufficiently geographically separated from the original data center. Implementing geo-replication is expensive as every data update operation in the original data center must be replicated in the backup. Running the application and the replication service in parallel is cost effective but creates a trade-off between potential replication consistency and data loss and reduced application performance due to network resource contention. We model this trade-off and provide a control-theoretical solution based on Model Predictive Control to dynamically allocate network bandwidth to accommodate the objectives of both replication and application data streams. We evaluate our control solution through simulations emulating the individual services, their traffic flows, and the shared network resource. The MPC solution is able to maintain a consistent performance over periods of persistent overload, and is quickly able to indiscriminately recover once the system return to a stable state. Additionally, the MPC balances the two objectives of consistency and performance according to the proportions specified in the objective function.

Interference between co-located VMs may lead to performance fluctuations and degradation, especially in overbooked datacenters. To limit this problem, VMs access to physical resources needs to be controlled to ensure certain degree of isolation among them. However, the mapping between virtual and physical resources must be performed in a dynamic way so that it can be adapted to the changing applications requirements, as well as to the different set of co-located VMs. To address this problem we propose a twofold approach: (1) a Quality of Service (QoS) scheme that provides different isolation levels for VMs with different QoS requirements, and (2) a self-adaptive fuzzy Q-learning capacity controller that proactively readjusts the isolation degree based on applications performance. Our evaluation based on real cloud applications and workloads demonstrates that the efficient, adaptive mapping between VMs and physical resources reduces the interference between VMs, enabling the possibility of co-locating more VMs, increases overall utilization, and ensures the performance of critical applications while providing more resources to the low QoS applications.

Current open issues regarding cloud computing include the support for nontrivial Quality of Service-related Service Level Objectives (SLOs) and reducing the energy footprint of data centers. One strategy that can contribute to both is the integration of accelerators as specialized resources within the cloud system. In particular, Field Programmable Gate Arrays (FPGAs) exhibit an excellent performance/energy consumption ratio that can be harnessed to achieve these goals. In this paper, a multilevel cloud scheduling framework is described, and several FPGA-aware node level scheduling strategies (applied at the hypervisor level) are explored and analyzed. These strategies are based on the use of a multiobjective metric aimed at providing Quality of Service (QoS) support. Results show how the proposed FPGA-aware scheduling policies increment the number of users requests serviced with their SLOs fulfilled while energy consumption is minimized. In particular, evaluation results of a use case based on a multimedia application show that the proposal can save more than 20% of the total energy compared with other baseline algorithms while a higher percentage of Service Level Agreement (SLA) is fulfilled.

Cloud computing and virtualized infrastructures are currently the baseline environments for the provision of services in different application domains. While the number of service consumers increasingly grows, service providers aim at exploiting infrastructures that enable non-disruptive service provisioning, thus minimizing or even eliminating downtime. Nonetheless, to achieve the latter current approaches are either application-specific or cost inefficient, requiring the use of dedicated hardware. In this paper we present the reference architecture of a fault-tolerance scheme, which not only enhances cloud environments with the aforementioned capabilities but also achieves high-performance as required by mission critical every day applications. To realize the proposed approach, a new paradigm for memory and I/O externalization and consolidation is introduced, while current implementation references are also provided.