Mathias Bjorkqvist

Opportunistic Service Provisioning in the Cloud

There is an emerging trend to deploy services in cloud environments due to their flexibility in providing virtual capacity and pay-as-you-go billing features. Cost-aware services demand computation capacity such as virtual machines (VMs) from a cloud operator according to the workload (i.e., service invocations) and pay for the amount of capacity used following billing contracts. However, as recent empirical studies show, the performance variability, i.e., non-uniform VM performance, is inherently higher than in private hosting platforms, since cloud platforms provide VMs running on top of typically heterogeneous hardware shared by multiple clients. Consequently, the provisioning of service capacity in a cloud needs to consider workload variability as well as varying VM performance. We propose an opportunistic service replication policy that leverages the variability in VM performance, as well as the on-demand billing features of the cloud. Our objective is to minimize the service provisioning costs by keeping a lower number of faster VMs, while maintaining target system utilization. Our evaluation results on traces collected from in-production systems show that the proposed policy achieves significant cost savings and low response times.

Minimizing retrieval latency for content cloud

Content cloud systems, e.g. CloudFront [1] and CloudBurst [2], in which content items are retrieved by end-users from the edge nodes of the cloud, are becoming increasingly popular. The retrieval latency in content clouds depends on content availability in the edge nodes, which in turn depends on the caching policy at the edge nodes. In case of local content unavailability (i.e., a cache miss), edge nodes resort to source selection strategies to retrieve the content items either vertically from the central server, or horizontally from other edge nodes. Consequently, managing the latency in content clouds needs to take into account several interrelated issues: asymmetric bandwidth and caching capacity for both source types as well as edge node heterogeneity in terms of caching policies and source selection strategies applied. In this paper, we study the problem of minimizing the retrieval latency considering both caching and retrieval capacity of the edge nodes and server simultaneously. We derive analytical models to evaluate the content retrieval latency under two source selection strategies, i.e., Random and Shortest-Queue, and three caching policies: selfish, collective, and a novel caching policy that we call the adaptive caching policy. Our analysis allows the quantification of the interrelated performance impacts of caching and retrieval capacity and the exploration of the corresponding design space. In particular, we show that the adaptive caching policy combined with Shortest-Queue selection scales well with various network configurations and adapts to the load changes in our simulation and analytical results.

Dynamic Replication in Service-Oriented Systems

Service-oriented systems, consisting of atomic services and their compositions hosted in service composition execution engines (CEEs), are commonly deployed to deliver web applications. As the workloads of applications fluctuate over time, it is economical to autonomously and dynamically adjust system capacity, i.e., the number of replicas for atomic services and CEEs. In this paper, we propose a novel replica provisioning policy, Resos, which adjusts the number of CEE and service replicas periodically based on the predicted workloads such that all replicas are well utilized at the target values. In particular, Resos models the workload balance and dependency between CEE and service replicas by estimating the probability that threads of CEE replicas are not blocked by I/O. Moreover, we derive the analytical bounds of CEE effective utilization and illustrate the cause of low nominal utilization at CEE replicas. We evaluate Resos on a simulated service-oriented system, which hosts CEE and service replicas on multi-threaded servers. The evaluated workload is derived from utilization traces collected from production systems. Through simulation, we demonstrate that Resos effectively reduces the number of required replicas while maintaining target utilization and lowering the response times of requests.

Cost-driven service provisioning in hybrid clouds

Hybrid clouds, which comprise nodes both in a private cloud and in a public cloud, have emerged as a new model for service providers to deploy their services. However, given Quality-of-Service requirements for each service, the question of on how many private and public nodes to deploy the services in the most cost-effective way remains to be answered. The challenges faced in the hybrid cloud stem from the disparate time-varying requests across multiple services, the different cost structures of both types of nodes, and the performance characteristics of nodes. In this paper, we propose a novel algorithm to dynamically optimize the allocation of private and public nodes across services, with special focus on the performance-cost tradeoff between private and public nodes. The algorithm is based on an analytical cost-performance framework for service deployment in hybrid clouds. Our evaluation results based on trace-driven simulation show that our proposed node allocation algorithm can effectively achieve a good cost-performance ratio, compared to the deployment of purely public and private cloud.

QoS-Aware Service VM Provisioning in Clouds: Experiences, Models, and Cost Analysis

Recent studies show that service systems hosted in clouds can elastically scale the provisioning of pre-configured virtual machines VMs with workload demands, but suffer from performance variability, particularly from varying response times. Service management in clouds is further complicated especially when aiming at striking an optimal trade-off between cost i.e., proportional to the number and types of VM instances and the fulfillment of quality-of-service QoS properties e.g., a system should serve at least 30i¾?requests per second for more than 90% of the time. In this paper, we develop a QoS-aware VM provisioning policy for service systems in clouds with high capacity variability, using experimental as well as modeling approaches. Using a wiki service hosted in a private cloud, we empirically quantify the QoS variability of a single VM with different configurations in terms of capacity. We develop a Markovian framework which explicitly models the capacity variability of a service cluster and derives a probability distribution of QoS fulfillment. To achieve the guaranteed QoS at minimal cost, we construct theoretical and numerical cost analyses, which facilitate the search for an optimal size of a given VM configuration, and additionally support the comparison between VM configurations.

Design and implementation of a key-lifecycle management system

Key management is the Achilles’ heel of cryptography. This work presents a novel Key-Lifecycle Management System (KLMS), which addresses two issues that have not been addressed comprehensively so far. First, KLMS introduces a pattern-based method to simplify and to automate the deployment task for keys and certificates, i.e., the task of associating them with endpoints that use them. Currently, the best practice is often a manual process, which does not scale and suffers from human error. Our approach eliminates these problems and specifically takes into account the lifecycle of keys and certificates. The result is a centralized, scalable system, addressing the current demand for automation of key management. Second, KLMS provides a novel form of strict access control to keys and realizes the first cryptographically sound and secure access-control policy for a key-management interface. Strict access control takes into account the cryptographic semantics of certain key-management operations (such as key wrapping and key derivation) to prevent attacks through the interface, which plagued earlier key-management interfaces with less sophisticated access control. Moreover, KLMS addresses the needs of a variety of different applications and endpoints, and includes an interface to the Key Management Interoperability Protocol (KMIP) that is currently under standardization.

Optimizing service replication in clouds

The load on todays service-oriented systems is strongly varying in time. It is advantageous to conserve energy by adapting the number of replicas according to the recent load. Over-provisioning of service replicas is to be avoided, since it increases the operating costs. Under-provisioning of service replicas leads to serious performance degradation and violates service-level agreements. To reduce energy consumption and maintain appropriate performance, we study two service replication strategies: (1) arrival rate based and (2) response time based policy. By simulation, we show that the average number of service replicas and response time can be reduced especially when combining our proposed replication strategies and load balancing schemes.

Load-Balancing Dynamic Service Binding in Composition Execution Engines

Performance and scalability of service-oriented applications, such as Web service compositions or business processes, depend on the dynamically bound services. In order to handle an increasing number of clients, load-balancing techniques are important. In this paper we assume the presence of multiple functionally equivalent services and explore different load-balancing algorithms to dynamically select service bindings with the goal to reduce average service response time. Using mathematical queueing models of service performance and simulation, we compare different service selection algorithms, including Static Lottery, Round-Robin, and Shortest-Queue. Furthermore, we propose linear and quadratic Dynamic Lottery service selection algorithms, which assign and periodically update service selection probabilities according to monitored average service response time. Our simulation environment models both stateless and stateful services and offers a wide range of service performance models with different degrees in the variation of service response time. While the Shortest-Queue algorithm performs best in simulation settings with only stateless services or low variance of service response time, the Round-Robin and Dynamic Lottery algorithms work best in settings with stateful services and high variance of service performance.

Content Retrieval Delay Driven by Caching Policy and Source Selection

In this paper we study the content retrieval delay in a hybrid content distribution system, e.g., emerging content clouds [1], where a requested content item can be vertically retrieved from the central server and horizontally retrieved from network nodes. The content retrieval delay depends on the load intensities of the retrieval sources, which have asymmetric system properties such as bandwidth and cache capacity. The retrieval traffic arises due to heterogeneous content availability, i.e., content diffusion resulting from the applied caching policies, and the selection of retrieval sources. To optimize the retrieval delay, the advantages of the network nodes should be utilized while also leveraging the caching and retrieval capacity of the server. The traffic loads and latency of a given combination of source selection and caching policy is derived based on the content diffusion and distribution in the entire system. The simulation and analytical results show that satisfactory content retrieval delay is achieved when the retrieval selection is load aware and the caching policies can effectively utilize the cache storage and retrieval capacity of both the network nodes and the server.

Power of redundancy: Designing partial replication for multi-tier applications

Replicating redundant requests has been shown to be an effective mechanism to defend application performance from high capacity variability — the common pitfall in the cloud. While the prior art centers on single-tier systems, it still remains an open question how to design replication strategies for distributed multi-tier systems, where interference from neighboring workloads is entangled with complex tier interdependency. In this paper, we design a first of its kind PArtial REplication system, sPARE, that replicates and dispatches read-only workloads for multi-tier web applications, determining replication factors per tier. The two key components of sPARE are (i) the variability-aware replicator that coordinates the replication levels on all tiers via an iterative searching algorithm, and (ii) the replication-aware arbiter that uses a novel token-based arbitration algorithm (TAD) to dispatch requests in each tier. We evaluate sPARE on web serving and web searching applications, i.e., MediaWiki and Solr, deployed on our private cloud testbed. Our results based on various interference patterns and traffic loads show that sPARE is able to improve the tail latency of MediaWiki and Solr by a factor of almost 2.7x and 2.9x, respectively.