Frank Griesinger

Cloud orchestration features: are tools fit for purpose?

Even though the cloud era has begun almost one decade ago, many problems of the first hour are still around. Vendor lock-in and poor tool support hinder users from taking full advantage of main cloud features: dynamic and scale. This has given rise to tools that target the seamless management and orchestration of cloud applications. All these tools promise similar capabilities and are barely distinguishable what makes it hard to select the right tool. In this paper, we objectively investigate required and desired features of such tools and give a definition of them. We then select three open-source tools (Brooklyn, Cloudify, Stratos) and compare them according to the features they support using our experience gained from deploying and operating a standard three-tier application. This exercise leads to a fine-grained feature list that enables the comparison of such tools based on objective criteria as well as a rating of three popular cloud orchestration tools. In addition, it leads to the insight that the tools are on the right track, but that further development and particularly research is necessary to satisfy all demands.

Axe: A Novel Approach for Generic, Flexible, and Comprehensive Monitoring and Adaptation of Cross-Cloud Applications

The vendor lock-in has been a major problem since cloud computing has evolved as on the one hand side hinders a quick transition between cloud providers and at the other hand side hinders an application deployment over various clouds at the same time (cross-cloud deployment). While the rise of cross-cloud deployment tools has to some extend limited the impact of vendor lock-in and given more freedom to operators, the fact that applications now are spread out over more than one cloud platform tremendously complicates matters: Either the operator has to interact with the interfaces of various cloud providers or he has to apply custom management tools. This is particularly true when it comes to the task of auto-scaling an application and adapting it to load changes. This paper introduces a novel approach to monitoring and adaptation management that is able to flexibly gather various monitoring data from virtual machines distributed across cloud providers, to dynamically aggregate the data in the cheapest possible manner, and finally, to evaluate the processed data in order to adapt the application according to user-defined rules.

Experiences of models@run-time with EMF and CDO

Model-drivenengineering promotes models and modeltrans- formations as the primary assets in software development. The models@run-time approach provides an abstract rep- resentation of a system at run-time, whereby changes in the model and the system are constantly reflected on each other. In this paper, we report on more than three years of experience with realising models@run-time in scalable cloud scenarios using a technology stack consisting of the Eclipse Modelling Framework (EMF) and Connected Data Objects(CDO).We establish requirements for the three roles domain-specific language (DSL) designer, developer, and operator, and compare them against the capabilities of EM- F/CDO. It turns out that this technology stack is well-suited for DSL designers, but less recommendable for developers and even less suited for operators. For these roles, we experi- enced a steep learning curve and several lacking features that hinder the implementation of models@run-time in scalable cloud scenarios. Performance experiences show limitations for write heavy scenarios with an increasing amount of total elements. While we do not discourage the use of EMF/CDO for such scenarios, we recommend that its adoption for sim- ilar use cases is carefully evaluated until this technology stack has realised our wish list of advanced features.

Beyond Mere Application Structure Thoughts on the Future of Cloud Orchestration Tools

Abstract Managing cloud applications running on IaaS is complicated and error prone. This is why DevOps tools and application description languages have been emerging. While these tools and languages enable the user to define the application and communication structure based on application components, they lack the possibility to define sophisticated communication patterns including the wiring on instance level. This paper details these shortcomings and presents approaches to overcome them. In particular, they we propose (i) adding boundaries to wiring specifications and (ii) introducing a higher-level abstraction—called facet —on top of the application. The combination of both concepts allows specifying wiring on basis of logical units and their relations. Hence, the concepts overcome general wiring problems that currently exist in cloud orchestration tools. In addition to that, the introduction of facets improves the re-use of components across di ff erent applications.

A Toolkit Based Architecture for Optimizing Cloud Management, Performance Evaluation and Provider Selection Processes

Cloud environments are criticized for their volatility in performance aspects, making it extremely difficult for time- critical applications owners to perform the decisive step for migration and owners of SaaS to present performance vs cost tradeoffs to their customers when acting as IaaS customers. The aim of this work is to present an architectural approach based on which a)IaaS providers may enhance the stability and performance effectiveness of their infrastructures, through overhead modelling, runtime analysis and optimal groupings of concurrently running services, b) IaaS adopters to understand the application computational nature, investigate abstracted QoS metrics for providers ranking, minimize procurement time and selection processes, automate deployment/orchestration and monitor the maintenance of their SLA c) 3rd parties to act as independent validators of IaaS QoS features, through a constant monitoring and benchmarking process for performance and SLA evaluation.

A Cross-Layer BPaaS Adaptation Framework

The notion of a BPaaS is currently taking a momentum as many organisations attempt to move and offer their business processes (BPs) in the cloud. Such BPs need to be adaptively provisioned so as to sustain the service level promised in the respective SLA. However, current cloud-based adaptation frameworks cannot cover all possible abstraction levels and usually rely on simplistic adaptation rules. As such, this paper proposes a novel BPaaS adaptation framework able to orchestrate actions on different abstraction levels so as to better address the current problematic situation. This framework can support the dynamic generation of adaptation workflows as well as the recording of the adaptation history for analysis purposes. It is also coupled with the CAMEL language which has been extended to support the specification of cross-level adaptation workflows.

The cloud application modelling and execution language (CAMEL)

Cloud computing provides ubiquitous networked access to a shared and virtualised pool of computing capabilities that can be provisioned with minimal management effort [27]. Cloud applications are deployed on cloud infrastructures and delivered as services. The PaaSage project aims to facilitate the modelling and execution of cloud applications by leveraging model-driven engineering (MDE) and by exploiting multiple cloud infrastructures. The Cloud Application Modelling and Execution Language (CAMEL) is the core modelling and execution language developed in the PaaSage project and enables the specification of multiple aspects of cross-cloud applications (i.e., applications deployed across multiple private, public, or hybrid cloud infrastructures). By exploiting models at both designand run-time, and by allowing both direct and programmatic manipulation of models, CAMEL enables the management of self-adaptive cross-cloud applications (i.e., cross-cloud applications that autonomously adapt to changes in the environment, requirements, and usage). In this paper, we describe the design and implementation of CAMEL, with emphasis on the integration of heterogeneous domain-specific languages (DSLs) that cover different aspects of self-adaptive cross-cloud applications. Moreover, we provide a real-world running example to illustrate how to specify models in a concrete textual syntax and how to dynamically adapt these models during the application life cycle. Finally, we provide an evaluation of CAMEL’s usability and usefulness, based on the technology acceptance model (TAM).

A DMN-Based Approach for Dynamic Deployment Modelling of Cloud Applications

Cloud computing is well suited for applications with a distributed architecture and dynamic demand of resources. Yet, current approaches to model cloud application deployment do not cater for the application’s dynamic nature and its rapidly changing business requirements. The static description of deployments results in a lack of reusability and also lacks an integrated way to adapt to the current context. To reuse and refine the deployment model, we introduce a simple decision layer on top of a cloud application description, which abstracts from the actual deployment language and allows assembling the deployment model from existing model fragments. Those fragments are chosen based on the input of the decision process. We define an architecture for the decision layer and sketch an implementation based on CAMEL, DMN, and ADOxx. The benefits of the decision layer are illustrated by two use cases. Our approach shifts the focus from a static to a dynamic and reusable modelling process, which also reduces the modeller’s effort.