Orchestration Framework Research Articles

Code health in EOS: Improving test infrastructure and overall service quality

During the last few years, the EOS distributed storage system at CERN has seen a steady increase in use, both in terms of traffic volume as well as sheer amount of stored data. This has brought the unwelcome side effect of stretching the EOS software stack to its design constraints, resulting in frequent user-facing issues and occasional downtime of critical services. In this paper, we discuss the challenges of adapting the software to meet the increasing demands, while at the same time preserving functionality without breaking existing features or introducing new bugs. We document our efforts in modernizing and stabilizing the codebase, through the refactoring of legacy code, introduction of widespread unit testing, as well as leveraging Kubernetes to build a comprehensive test orchestration framework capable of stressing every aspect of an EOS installation, with the goal of discovering bottlenecks and instabilities before they reach production.

Climate Ambition and Sustainable Development for a New Decade: A Catalytic Framework

This paper examines the Global Climate Action Agenda (GCAA) and discusses options to improve sub- and non-state involvement in post-2020 climate governance. A framework that stimulates sub- and non-state action is a necessary complement to national governmental action, as the latter falls short of achieving low-carbon and climate-resilient development as envisaged in the Paris Agreement. Applying design principles for an ideal-type orchestration framework, we review literature and gather expert judgements to assess whether the GCAA has been collaborative, comprehensive, evaluative and catalytic. Results show that there has been greater collaboration and coordination among orchestrators, for instance, in the organization of events. However, mobilization efforts remain event-driven and too little effort is invested in understanding the progress of sub- and non-state action. Data collection has improved, although more sophisticated indicators are needed to evaluate climate and sustainable development impacts. Finally, the GCAA has recorded more action, but relatively little by actors in developing countries. As the world seeks to recover from the COVID-19 crisis, a post-2020 GCAA could make a vital contribution to climate action in challenging times by helping governments keep climate and enhance commitments; strengthening capacity for sub- and non-state action; enabling accountability; and helping to realize sustainable development.

GA-Par: Dependable Microservice Orchestration Framework for Geo-Distributed Clouds

Recent advances in composing Cloud applications have been driven by deployments of inter-networking heterogeneous microservices across multiple Cloud datacenters. System dependability has been of the upmost importance and criticality to both service vendors and customers. Security, a measurable attribute, is increasingly regarded as the representative example of dependability. Literally, with the increment of microservice types and dynamicity, applications are exposed to aggravated internal security threats and externally environmental uncertainties. Existing work mainly focuses on the QoS-aware composition of native VM-based Cloud application components, while ignoring uncertainties and security risks among interactive and interdependent container-based microservices. Still, orchestrating a set of microservices across datacenters under those constraints remains computationally intractable. This paper describes a new dependable microservice orchestration framework GA-Par to effectively select and deploy microservices whilst reducing the discrepancy between user security requirements and actual service provision. We adopt a hybrid (both whitebox and blackbox based) approach to measure the satisfaction of security requirement and the environmental impact of network QoS on system dependability. Due to the exponential grow of solution space, we develop a parallel Genetic Algorithm framework based on Spark to accelerate the operations for calculating the optimal or near-optimal solution. Large-scale real world datasets are utilized to validate models and orchestration approach. Experiments show that our solution outperforms the greedy-based security aware method with 42.34 percent improvement. GA-Par is roughly 4x faster than a Hadoop-based genetic algorithm solver and the effectiveness can be constantly guaranteed under different application scales.

Intelligent VNF Orchestration and Flow Scheduling via Model-Assisted Deep Reinforcement Learning

Hosting virtualized network functions (VNF) has been regarded as an effective way to realize network function virtualization (NFV). Considering the cost diversity in cloud computing, from the perspective of service providers, it is significant to orchestrate the VNFs and schedule the traffic flows for network utility maximization (NUM) as it implies maximal revenue. However, traditional heuristic solutions based on optimization models usually follow some assumptions, limiting their applicability. Recent studies have shown that deep reinforcement learning (DRL) is a promising way to tackle such limitations. However, DRL agent training also suffers from slow convergence problem, especially with complex control problems. We notice that optimization models actually can be applied to accelerate the DRL training. Therefore, we are motivated to design a model-assisted DRL framework for VNF orchestration in this paper. Other than letting the agent blindly explore actions, the heuristic solutions are used to guide the training process. Based on such principle, the DRL framework is also redesigned accordingly. Experiment results validate the high efficiency of our model-assisted DRL framework as it not only converges 23× faster than traditional DRL algorithm, but also with higher performance at the same time.

An Artificial Intelligence Framework for Slice Deployment and Orchestration in 5G Networks

Network slicing is a key enabler to successfully support 5G services with specific requirements and priorities. Due to the diversity of these services, slice deployment and orchestration are essential to guarantee service performance in a cost-effective way. Here, we propose an Artificial Intelligence framework for cross-slice admission and congestion control that simultaneously considers communication, computing, and storage resources to maximize resources utilization and operator revenue. First, we propose a smart feature extraction solution to analyze the characteristics of incoming requests together with the already deployed slices, and then automatically evaluates the request requirements to make appropriate decisions. Second, we design an online algorithm that controls the slice admission based on their priorities, the arrival and departure characteristics, and the available resources. To mitigate system overloading, our framework dynamically adjusts resources allocated to low priority slices, thereby reducing the dropping probability of new slice requests. The proposed algorithm offers outstanding advantages over traditional static approaches by automatically adapting the controller decisions to the system changes. Simulation results show that our framework significantly improves the resource utilization and reduces the slice request dropping probabilities up to 44% as compared to the baseline schemes.

Service Shifting: A Paradigm for Service Resilience in 5G

Many real-world services can be provided through multiple virtual network function (VNF) graphs, corresponding, for example, to high- and low-complexity variants of the service itself. Based on this observation, we extend the concept of service scaling in network orchestration to service shifting, that is, upgrading or downgrading the VNF graph to use among those implementing the same service. Service shifting can serve multiple goals, from reducing operational costs to reacting to infrastructure problems. Furthermore, it enhances the flexibility of service-level agreements between network operators and third-party content providers (verticals). In this article, we introduce and describe the service shifting concept, its benefits, and the associated challenges, with special reference to how service shifting can be integrated within real-world 5G architectures and implementations. We conclude that existing network orchestration frameworks can be easily extended to support service shifting, and its adoption has the potential to make 5G network slices easier for the operators to manage under highload conditions, while still meeting the verticals' requirements.

Interconnecting networks with optimized service provisioning

A recent trend of peering at geo-diversified Internet exchange points (IXPs) has empowered decades-old proposal of inter-networking and opened up new avenues of business ventures. IP-transit, cloud direct and remote peering are a few important amongst numerous proposals of service provisioning capitalizing on this peering infrastructure support across domains. Enduring these business proposals becomes a challenging task, especially when the increased dependency of enterprises over the Internet is affirmed. Volatile traffic priorities necessitate different strategies of flow management for each pattern of enterprise traffic. Providing diverse service guarantees to each traffic class require careful selection of resource allocation and compliance of inter-domain policies. In this paper, we propose a novel orchestration framework that helps to stitch end-to-end traffic engineering compliant multiple paths. The framework enables prioritized management of various traffic classes in a centralized manner by employing software defined networking paradigm. Abstraction of multi-graph from the inter-connectivity of peering anchors helps to gear service provisioning spanning across multiple domains. Beside presenting details of our framework, we have articulated use cases highlighting the efficacy of our proposal. We have observed a maximum increase of 26.52% in throughput using proposed model compared with an optimization formulation from literature. Our results imply transparent utility of this formulation for various network topologies and traffic loads.

Online Orchestration of Cross-Edge Service Function Chaining for Cost-Efficient Edge Computing

Edge computing (EC) has quickly ascended to be the de-facto standard for hosting emerging low-latency applications, as exemplified by intelligent video surveillance, Internet of Vehicles, and augmented reality. For EC, service function chaining is envisioned as a promising approach to configure various services in an agile, flexible, and cost-efficient manner. When running on top of geographically dispersed edge clouds, fully unleashing the benefits of service function chaining is, however, by no means trivial. In this paper, we propose an online orchestration framework for cross-edge service function chaining, which aims to maximize the holistic cost efficiency, via jointly optimizing the resource provisioning and traffic routing on-the-fly. This long-term cost minimization problem is difficult since it is NP-hard and involves future uncertain information. To simultaneously address these dual challenges, we carefully combine an online optimization technique with an approximate optimization method in a joint optimization framework, through: 1) decomposing the long-term problem into a series of one-shot fractional problem with a regularization technique and 2) rounding the fractional solution to a near-optimal integral solution with a randomized dependent scheme that preserves the solution feasibility. The resulting online algorithm achieves an outstanding performance guarantee, as verified by both rigorous theoretical analysis and extensive trace-driven simulations.

An Optimization-Enhanced MANO for Energy-Efficient 5G Networks

5G network nodes, fronthaul and backhaul alike, will have both forwarding and computational capabilities. This makes energy-efficient network management more challenging, as decisions, such as activating or deactivating a node, impact on both the ability of the network to route traffic and the amount of processing it can perform. To this end, we formulate an optimization problem accounting for the main features of 5G nodes and the traffic they serve, allowing joint decisions about: 1) the nodes to activate; 2) the network functions they run; and 3) the traffic routing. Our optimization module is integrated within the management and orchestration framework of 5G, thus enabling swift and high-quality decisions. We test our scheme with both a real-world testbed based on OpenStack and OpenDaylight, and a large-scale emulated network whose topology and traffic come from a real-world mobile operator, finding it to consistently outperform state-of-the art alternatives and closely match the optimum.

BCDN: A proof of concept model for blockchain-aided CDN orchestration and routing

As the conventional definition, the Content Delivery Networks (CDN) build surrogate servers on top of the Internet architecture to provide contents to the end users as fast as possible. However, the high client requests lead to overloading of the surrogate servers. Despite the routing procedures which try to distribute these high requests among end users, routing these requests to other surrogate servers also emerges as a crucial content provisioning problem. This leads another challenge, which is the high latency that is an inevitable consequence of the conventional CDN architecture. To this end, it is necessary to design a CDN architecture which provides orchestration framework for virtual instances and routing procedure among surrogate servers, by considering the client trends. Moreover, during the routing procedure, the content provisioning should also be taken into account to prevent latency arisen from caching problems. For this purpose, we propose a proof of concept CDN model which combine a novel orchestration framework and blockchain architecture, Blockchain-Aided CDN, namely BCDN. Here the main reason to use Blockchain methodology is its specific features of distributed contracts and fully automated reliable flow generation. With this proposed orchestration framework, we provide dynamically changing virtual instances of the surrogate servers according to client requests and resource capacity. Moreover, by using blockchain network, we ensure that the our proposed CDN model can handle with evenly distributed high bandwidth requests by considering status of other Points of Present (PoPs). Also, deducing from the blockchain data enable surrogate servers to make content provisioning.

An Orchestration Framework for IoT Devices Based on Public Key Infrastructure (PKI)

An Orchestration Framework for IoT Devices Based on Public Key Infrastructure (PKI)

A Comprehensive Feature Comparison Study of Open-Source Container Orchestration Frameworks

(1) Background: Container orchestration frameworks provide support for management of complex distributed applications. Different frameworks have emerged only recently, and they have been in constant evolution as new features are being introduced. This reality makes it difficult for practitioners and researchers to maintain a clear view of the technology space. (2) Methods: we present a descriptive feature comparison study of the three most prominent orchestration frameworks: Docker Swarm, Kubernetes, and Mesos, which can be combined with Marathon, Aurora or DC/OS. This study aims at (i) identifying the common and unique features of all frameworks, (ii) comparing these frameworks qualitatively and quantitatively with respect to genericity in terms of supported features, and (iii) investigating the maturity and stability of the frameworks as well as the pioneering nature of each framework by studying the historical evolution of the frameworks on GitHub. (3) Results: (i) we have identified 124 common features and 54 unique features that we divided into a taxonomy of 9 functional aspects and 27 functional sub-aspects. (ii) Kubernetes supports the highest number of accumulated common and unique features for all 9 functional aspects; however, no evidence has been found for significant differences in genericity with Docker Swarm and DC/OS. (iii) Very little feature deprecations have been found and 15 out of 27 sub-aspects have been identified as mature and stable. These are pioneered in descending order by Kubernetes, Mesos, and Marathon. (4) Conclusion: there is a broad and mature foundation that underpins all container orchestration frameworks. Likely areas for further evolution and innovation include system support for improved cluster security and container security, performance isolation of GPU, disk and network resources, and network plugin architectures.

A scalable network-aware framework for cloud monitoring orchestration

Monitoring components are implanted in clouds to evaluate performance, detect the failures, and assess component interactions by message analysis. In this paper, we propose Monitoring as a Service (MaaS) installed across its software defined network in the cloud. All switches in datacenter only forward traffic. Specific SDN application on top of the network controller has been implemented in order to orchestrate multiple network tenants monitoring needs. Applications declare the required observation for traffics and their specific monitoring needs to the Maas. Therefore, a set of virtual monitoring functions (vMF) are prescribed to be placed in datacenter for flows. An optimal placement algorithm places vMF with respect to the network and computing utilization maximization objectives. Network objective refers to minimize traffic delay for flows needed to be monitored and computing objective expresses balancing nodes computing resources. The optimal placement of virtual network functions is known to be an NP-hard problem. Compared to the existing work, we discovered different problem which how a vMF can be split into smaller pieces to decrease the total placement cost for a set of flows required, based on four patterns: free, parser-collocation, job-collocation, and full-collocation. Moreover, we proposed three heuristics to make our placement algorithm scalable for a large network, called, chain partitioning, topology partitioning, and zoning. We show the feasibility of our approach in a large dataset consists of 540k nodes and 11.5M edges, with about 40 requests (flows) at the same time. Furthermore, the proposed solution saves at least 20 percent of total monitoring cost, in average, compared to the latest related work.

An online machine learning framework for early detection of product failures in an Industry 4.0 context

ABSTRACTCurrent paradigms such as the Internet of Things (IoT) and cyber-physical systems are transforming production environments, where related processes are not only faster and with higher standards, but also more flexible and adaptable to changes in the environment. To address the ever-increasing flexibility requirements while keeping current production standards, a new set of technologies is needed. This paper presents an IoT machine learning and orchestration framework, applied to detection of failures of surface mount devices during production. The paper shows how to build a scalable and flexible system for real-time, online machine learning. Furthermore, the approach is evaluated by using a novel and realistic simulation of a production line for electronic devices as a case study. The system evaluation is done in a holistic manner by analyzing various aspects involving the software architecture, computational scalability, model accuracy, production performance, among others.

An orchestration framework for monitoring and analysis of container-based logs and systems

Container technology provides a rapid way of running applications and hence draws it attention to recent times. Container applications are smaller in size and much flexible in management compared to traditional virtual machines. Since containers are non-persistence, the complexity of application logs also increases with number of deployments. Even with wide range of log management tools available, it requires a combination of few for providing a complete logging solution. Thus, it requires a framework which could orchestrate and identify the status of each integrated tools for providing smooth functioning across the entire log system. In this paper, we discussed about the implementation of our orchestration framework, configuration details for communication across log systems and case study.

Topology-Aware Resource-Efficient Placement for High Availability Clusters Over Geo-Distributed Cloud Infrastructure

A management and orchestration framework (MANO) in network function virtualization (NFV) enables the agile deployment and operation of virtual network functions over a geographically distributed cloud infrastructure. This facilitates the deployment of redundancy models (i.e., high availability clusters) over different cloud centers, to guarantee the high availability of network services. In particular, in the telecommunications field, availability and resiliency are always required at a high level. Existing placement algorithms only consider one type of redundancy model at a given time. However, in reality, different redundancy configurations can be utilized to ensure the availability of virtual functions. In this article, we present an optimization model and topology-aware resource-efficient placement algorithm (TARE), which can be employed to optimally deploy high availability clusters with different redundancy configurations over geo-distributed cloud infrastructures. This model takes into account the different requirements of various high availability clusters in terms of bandwidth and computing resource demands. By simulation, the TARE has better performance than other baseline solutions in terms of the bandwidth usage, while maintaining an acceptable level of availability.

An orchestration framework for monitoring and analysis of container-based logs and systems

Container technology provides a rapid way of running applications and hence draws it attention to recent times. Container applications are smaller in size and much flexible in management compared to traditional virtual machines. Since containers are non-persistence, the complexity of application logs also increases with number of deployments. Even with wide range of log management tools available, it requires a combination of few for providing a complete logging solution. Thus, it requires a framework which could orchestrate and identify the status of each integrated tools for providing smooth functioning across the entire log system. In this paper, we discussed about the implementation of our orchestration framework, configuration details for communication across log systems and case study.

PhenoMeNal: processing and analysis of metabolomics data in the cloud.

BackgroundMetabolomics is the comprehensive study of a multitude of small molecules to gain insight into an organism's metabolism. The research field is dynamic and expanding with applications across biomedical, biotechnological, and many other applied biological domains. Its computationally intensive nature has driven requirements for open data formats, data repositories, and data analysis tools. However, the rapid progress has resulted in a mosaic of independent, and sometimes incompatible, analysis methods that are difficult to connect into a useful and complete data analysis solution.FindingsPhenoMeNal (Phenome and Metabolome aNalysis) is an advanced and complete solution to set up Infrastructure-as-a-Service (IaaS) that brings workflow-oriented, interoperable metabolomics data analysis platforms into the cloud. PhenoMeNal seamlessly integrates a wide array of existing open-source tools that are tested and packaged as Docker containers through the project's continuous integration process and deployed based on a kubernetes orchestration framework. It also provides a number of standardized, automated, and published analysis workflows in the user interfaces Galaxy, Jupyter, Luigi, and Pachyderm.ConclusionsPhenoMeNal constitutes a keystone solution in cloud e-infrastructures available for metabolomics. PhenoMeNal is a unique and complete solution for setting up cloud e-infrastructures through easy-to-use web interfaces that can be scaled to any custom public and private cloud environment. By harmonizing and automating software installation and configuration and through ready-to-use scientific workflow user interfaces, PhenoMeNal has succeeded in providing scientists with workflow-driven, reproducible, and shareable metabolomics data analysis platforms that are interfaced through standard data formats, representative datasets, versioned, and have been tested for reproducibility and interoperability. The elastic implementation of PhenoMeNal further allows easy adaptation of the infrastructure to other application areas and ‘omics research domains.

Cost-Efficient Orchestration of Containers in Clouds: A Vision, Architectural Elements, and Future Directions

This paper proposes an architectural framework for the efficient orchestration of containers in cloud environments. It centres around resource scheduling and rescheduling policies as well as autoscaling algorithms that enable the creation of elastic virtual clusters. In this way, the proposed framework enables the sharing of a computing environment between differing client applications packaged in containers, including web services, offline analytics jobs, and backend pre-processing tasks. The devised resource management algorithms and policies will improve utilization of the available virtual resources to reduce operational cost for the provider while satisfying the resource needs of various types of applications. The proposed algorithms will take factors that are previously omitted by other solutions into consideration, including 1) the pricing models of the acquired resources, 2) and the fault-tolerability of the applications, and 3) the QoS requirements of the running applications, such as the latencies and throughputs of the web services and the deadline of the analytical and pre-processing jobs. The proposed solutions will be evaluated by developing a prototype platform based on one of the existing container orchestration platforms.

Unicorn: Unified resource orchestration for multi-domain, geo-distributed data analytics

As the data volume increases exponentially over time, data-intensive analytics benefits substantially from multi-organizational, geographically-distributed, collaborative computing, where different organizations contribute various yet scarce resources, e.g., computation, storage and networking resources, to collaboratively collect, share and analyze extremely large amounts of data. By analyzing the data analytics trace from the Compact Muon Solenoid (CMS) experiment, one of the largest scientific experiments in the world, and systematically examining the design of existing resource management systems for clusters, we show that the multi-domain, geo-distributed, resource-disaggregated nature of this new paradigm calls for a framework to manage a large set of distributively-owned, heterogeneous resources, with the objective of efficient resource utilization, following the autonomy and privacy of different domains, and that the fundamental challenge for designing such a framework is: how to accurately discover and represent resource availability of a large set of distributively-owned, heterogeneous resources across different domains with minimal information exposure from each domain? Existing resource management systems are designed for single-domain clusters and cannot address this challenge. In this paper, we design Unicorn, the first unified resource orchestration framework for multi-domain, geo-distributed data analytics. In Unicorn, we encode the resource availability for each domain into resource state abstraction, a variant of the network view abstraction extended to accurately represent the availability of multiple resources with minimal information exposure using a set of linear inequalities. We then design a novel, efficient cross-domain query algorithm and a privacy-preserving resource information integration protocol to discover and integrate the accurate, minimal resource availability information for a set of data analytics jobs across different domains. In addition, Unicorn also contains a global resource orchestrator that computes optimal resource allocation decisions for data analytics jobs. We implement a prototype of Unicorn and present preliminary evaluation results to demonstrate its efficiency and efficacy. We also give a full demonstration of the Unicorn system at SuperComputing 2017.