CYCLONE in a Nutshell

Complete & Dynamic Multi-cloud Application Management
CYCLONE in a Nutshell
This process will also allow CYCLONE to measure the
CYCLONE is a Horizon 2020 innovation action funded technical progress of the project through tickets attached
by the European Commission which aims at to work items for particular development iteration
integrating existing cloud management software to (“sprint”) along with progress summaries of each sprint.
allow a unified management of federated clouds.
Application service providers (ASPs) develop, deploy,
and maintain complex computing platforms within
multiple cloud infrastructures to improve resilience,
responsiveness and elasticity of their applications.
The CYCLONE project targets the ASPs, providing
them with software and tools that
 Facilitate the deployment, management, and
applications, and
 Enhance the end-to-end security and
network management of those applications.
The CYCLONE project integrates and improves
existing, production-quality tools to achieve this aim.
CYCLONE focus areas and relevant work packages
The project partners have identified two flagship The main objectives of the CYCLONE project are:
domains: academic use cases for bioinformatics
 Improve cloud services in the Infrastructure-as-aresearch and use cases for a commercial platform for
Service (IaaS) layer by integrating network services
future energy management. These will guide the
into the cloud offering, allowing direct control over
initial development of the tools. Additional use cases,
virtual machine (VM) network accessibility, intraselected during the course of the project, will
site data access, and inter-site data transfers.
highlight missing, critical features and guide the
Develop tools that provide enhanced functionality
evolution of the software.
for cloud providers that agree to federate their
resources, such as dynamic allocation of bandwidth
The project adopts agile software processes, notably
SCRUM, to ensure that:
authentication mechanisms.
 The project delivers production quality code
 Provide tools that allow application developers to
that is thoroughly and systematically tested.
take advantage of features like VM coordination
 The needs of the real-world use cases are
within deployments, automated placement of
well covered by the CYCLONE features.
service components, and scaling of service
 Features for deploying and maintaining
components, essentially providing them with the
running cloud applications are provided.
means to develop a Platform-as-a-Service (PaaS)
 The project can respond to new
 Provide software that allows developers to ensure
development effort.
the end-to-end, secure use of data within their
 Constantly maintains a working version of the
application as well as secured access to remote
CYCLONE tools and platform.
data sources.
Demonstrate that the CYCLONE software meets the
The CYCLONE distributed testbed will play a critical
needs of concrete academic and commercial use
role, enabling the full, continuous validation of the
cases, while providing frequent, production-quality
releases of that software.
deployments on a federated bioinformatics cloud.
Manage Your Cloud Application Resources
Charles Loomis
Chief Technical Officer (SixSq)
CYCLONE Technical Coordinator and
WP6 Leader
As an application service provider, you must optimize the performance/cost ratio for your application by wisely
allocating cloud resources. Long-running services make this challenge more daunting because the application must
be re-optimized frequently to respond to the inevitable peaks and troughs in demand.
SlipStream, a core component of the CYCLONE platform developed by SixSq, already allows you to perform rather
sophisticated resource management for your cloud applications. Based on your application definition, SlipStream
Automatically provision all resources necessary to deploy the application,
Coordinate the configuration of the deployed application components, and
Scale an application based on your explicit requests.
By concentrating on applications, rather than virtual machines, SlipStream already significantly lightens your
management burden. CYCLONE will further lighten this burden by enhancing SlipStream with advanced brokering,
monitoring, and matchmaking features.
Currently with SlipStream, you choose the placement of your application components explicitly; each application
deployment requires human intervention. For most applications however, the placement constraints for an
application can be expressed as a static “policy”, opening up the possibility of automated placement. To do this, two
interconnected CYCLONE components will be developed:
A brokering database that contains characteristics of cloud service offers. The database will include a wide
variety of information, such as pricing from Cloud Service Providers like Exoscale, security certifications from
organizations like the Cloud Security Alliance, benchmarks from sites like CloudHarmony and metrics from
the application itself. Having an open schema will allow all of these types of information to be included,
allowing you to create rich deployment policies. The implementations will use JSON-LD as an interchange
A matchmaker that compares an application deployment policy vs. available cloud service offers. The
implementations will take an application description and the application deployment policy to find
acceptable cloud service offers from the brokering database. The policy can also reference a ranking
algorithm (e.g. minimum price) to prioritize the acceptable offers. The resulting prioritized list of offers can
be used for automated placement as well as refined manual placement. The policies will initially be based on
DMTF’s CIMI filtering syntax and will evolve to meet the needs of the CYCLONE use cases.
To achieve true “1-click” deployments, CYCLONE will extend SlipStream to use these components. You will be able to
specify the deployment policy for each component of the cloud application. With this and the application definition,
SlipStream will use the Matchmaker to find acceptable placement options and use the results for fully-automated
Going a step further, application monitoring will be coupled with the automated provisioning to provide fullyautomated scaling for applications using the CYCLONE platform. With this you can define self-managing cloud
applications, eliminating much of the tedious work involved with cloud application management.
You can already use SlipStream and other components of the CYCLONE platform. If the advanced features interest
you, we’d like to talk with you about your use case. Get in touch!
Figure 1: CYCLONE components that automate cloud resource provisioning through SlipStream.
CYCLONE Bioinformatics Use Case Scenarios
Christophe Blanchet
Chief Technical Officer (CNRS IFB)
Bioinformatics deals with the collection and efficient analysis of biological data, particularly genomic information
from DNA sequencers. The capability of modern sequencers to produce terabytes of information coupled with low
pricing (less than US$ 1000 for a human genome), and causes a "data deluge" that is being experienced by
researchers in this field.
To analyse these data, the scientific and industrial community daily use bioinformatics software that is characterized
by a high degree of fragmentation: literally hundreds of different software packages are regularly used for scientific
analyses with an incompatible variety of dependencies and a broad range of resource requirements. For this reason,
the bioinformatics community has strongly embraced cloud computing with its ability to provide customized
execution environments and dynamic resource allocation.
The French Institute of Bioinformatics - IFB consists of a national hub, the IFB-core, and 34 bioinformatics platforms
(PF) grouped into 6 regional centres spanning the entire French territory. The IFB currently deployed two cloud
facilities on its own premises, one in IFB-core (Orsay, France) and another one in the platform IFB-GenOuest
(Rennes, France), and two others in collaborations with the local computing centres in Lille and Strasbourg. In the
next years, the IFB aims to deploy a federated cloud infrastructure over the regional PFs. This cloud infrastructure is
devoted to the French life science community, research and industry, with services for the management and analysis
of life science data.
Thus, the CYCLONE consortium has identified several concrete bioinformatics use cases that aim to address some
specific well-identified limitations. For example, regarding the key technical areas of the CYCLONE project, Cloud
Access Management through cloud proxies and Matchmaking, Brokering, and Mediation of Cloud Resources will
provide the IFB with features to access other cloud infrastructures than the current national IFB’s, to integrate the
future cloud infrastructures that will be deployed by the regional IFB’s platform, and also to access external cloud
infrastructures from academic or commercial providers. Naturally, the infrastructure users, for example biomedical
staff analysing human biomedical data or researchers analysing genome sequences, will benefit from these features
while accessing the system and carrying out their analyses. Two bioinformatics use cases are already deployed on
the CYCLONE infrastructure.
The first CYCLONE bioinformatics use case scenario – Securing Human Biomedical Data
This use case is a single virtual machine (VM) application requiring enhanced security features such as a trusted
federated authentication mode and a deployment done only on certified (by the French Health Ministry) cloud
infrastructure. The use case workflow to ensure data security is illustrated in Figure 2. The cloud appliance NGSUnicancer is developed by the bioinformatics platform of the Centre Léon Bérard, Lyon, France in the context of the
project NGS-Clinique (INCA - Institut National du Cancer). It provides a simple web interface to launch the biomedical
genomic analysis pipeline. The appliance was enhanced by the Federation Provider developed by Technische
Universität Berlin and is ready for on-demand deployment on the IFB’s core cloud infrastructure. The user deploys
the appliance NGS-Unicancer through the IFB’s web interface in “1-click” and uses the CYCLONE federation provider
to get access to the VM web interface based on its identity in the federation. The user then easily uploads relevant
data, runs the analysis and gets the results.
Figure 2. Functional schema of the use case “Securing human biomedical data”.
The second CYCLONE bioinformatics use case - Cloud Virtual Pipeline for Microbial Genomes Analysis
This use case (which workflow is illustrated in Figure 3) is developed by the platform IFB-MIGALE, Jouy-en-Josas,
France. The application requires several components: a user web interface, a relational postgreSQL database, and a
complete computing cluster with a master and several nodes to perform the data-intensive analyses. This
infrastructure already running in a classical static way on bare-metal servers in IFB-MIGALE premises was ported to
the cloud and extended with a “1-click” deployment feature by using SlipStream recipes. The image was exported
from the IFB’s cloud and registered in the StratusLab Marketplace. Afterwards, IFB-core wrote a deployment recipe
based on SlipStream that instantiates the complete application with all the required VMs on the CYCLONE
Figure 3 Functional schema of the use case “Cloud virtual pipeline for microbial genomes analysis”.
CYCLONE Energy Use Case Scenario
Doris Hacker
Head of Trustful Cloud Solutions (QSC AG)
In order to comply with the “20/20/20” climate change mitigation goals of the EU, the energy supply has to be
changed from fossil energy generation to the increasing usage of renewable energy sources. The integration of CO2
free, volatile and decentralized, renewable energy resources, into the energy system leads to a new approach for
energy management systems. To efficiently incorporate the huge number of future participants to the energy
system, the integration with new cloud-based ICT technologies is essential.
The first described use case coming from the energy sector (see Figure 4) has the focus on the geographically widely
distributed, volatile energy production combined to a Virtual Power Plant. The main idea of a Virtual Power Plant is
to integrate the small, distributed energy resources for the generation of renewable energy and to combine them to
one reliable power plant to bring the produced energy into the existing power system. The decentralized energy
production leads to the necessity to collect and aggregate measurement data of energy production in real time all
over the grid. Concerning the different energy components coming with their own smart metering technologies, the
data management has to deal with heterogeneous data.
Given the sensitivity of the energy system the ICT platform has to fulfill high standards in security, availability and
scalability. All these requirements for the data management platform lead to a distributed ICT environment with a
trustable and heterogeneous multi-cloud base.
Figure 4. CYCLONE Energy Use Case.
1. M. Slawik, Y. Demchenko, J. I. Aznar Baranda, R. Branchat, C. Loomis, O. Lodygensky, C. Blanchet. CYCLONE Unified
Deployment and Management of Federated, Multi-Cloud Applications. Proceedings of the 5th workshop on
Network Infrastructure Services as part of Cloud Computing (NetCloud 2015). December, 2015.
2. Y. Demchenko, C. Dumitru, R. Koning, C. de Laat, T. Matselyukh, S. Filiposka, M. de Vos, D. Arbel, D. Regvart, T.
Karaliotas, K. Baumann. Open Cloud eXchange (OCX): A Pivot for Intercloud Services Federation in Multi-provider
Cloud Market Environment. Proceedings of the 4th IEEE International Workshop on Cloud Computing Interclouds,
Multiclouds, Federations, and Interoperability (Intercloud 2015), pp. 472-479.
3. M. Slawik, B. I. Zilci, F. Knaack, A. Küpper. The Open Service Compendium: Business-pertinent Cloud Service
Discovery, Assessment, and Selection. Proceedings of the 12th International Conference on Economics of Grids,
Clouds, Systems and Services (GECON 2015). Springer. View online
4. B. I. Zilci, M. Slawik, A. Küpper. Cloud Service Matchmaking using Constraint Programming. Proceedings of the
24th IEEE International Conference on Enabling Technologies: Infrastructure for Collaborative Enterprises (WETICE
2015), pp. 63-68, 2015.
5. J. I. Aznar Baranda. Contribution to the position paper of Inter-cloud Challenges, Expectations and Issues Cluster
within the EUCloudClusters initiative.
6. Y. Demchenko, C. Loomis, E. Escalona, J. I. Aznar Baranda, M. Slawick, O. Lodygensky. CYCLONE: Multi-cloud
applications deployment and management platform and ecosystem. Poster, EGI2015 Conference, May 2015.
7. Y. Demchenko. CYCLONE: Multi-cloud applications deployment and management platform for research and
academic community. Poster, TNC15 Conference, 14-18 June 2015. Download poster
Please refer to the CYCLONE website for updated information on publications and dissemination/demonstration activities for
forthcoming events
Newsletter #1 – CONTENTS
CYCLONE in a nutshell
Manage Your Cloud Application Resources
Charles Loomis, SixSQ
CYCLONE Bioinformatics Use Case Scenarios
Christophe Blanchet, CNRS IFB
CYCLONE Energy Use Case Scenario
Doris Hacker, QSC AG
Web Site
Recent CYCLONE publications
Contact us
CYCLONE at a glance
Project coordinator
Contract number
Call Identifier
Funding scheme
EC Contribution
January 2015 –
December 2017
Innovation action
3.84 M€
2.84 M€
Advanced Cloud
Infrastructures and
Services (ICT-07-2014)
Matteo Biancani, Interoute S.p.A.
Tel: +39 (0)6 61 52 40 33
Fax: +39 (0)6 61 52 40 99
Technical coordinator
Eduard Escalona, i2CAT
The CYCLONE consortium
Interoute S.p.A. Italy
SixSq Sàrl, Switzerland
Fundacio Privada I2CAT, Internet I Innovació
Digital a Catalunya Spain
Technische Universität Berlin,
Centre National de la Recherche Scientifique
Universiteit van Amsterdam,
The Netherlands
QSC AG Germany