in Protomoteca





  • Flavia MarzanoAssessora a Roma Semplice Flavia Marzano, Comune di Roma

Rosario Mignacca

Community cloud and OpenStack

Poli Strategici Nazionali e Community cloud. Come gestire e condividere virtualmente le infrastrutture di eccellenza dell'amministrazione tramite la tecnologia Openstack.




Sala Busti

Session A


Five years of OpenStack at CERN

CERN, the European Organization for Nuclear Research, have been running a private OpenStack cloud for five years. Our cloud has grown to >300K cores, providing Virtual Machines, Block Storage, Container Clusters to more than 3000 users.
We will present details about the Cloud Service, its architecture and operational challenges, as well as our plans for expansion.

OpenStack & Kubernetes together at full power

SUSE OpenStack Cloud 8 comes with two options for Kubernetes deployment:
OpenStack Magnum that let users to bring their own image, and preconfigured Heat templates to deploy ready to run SUSE CaaS Platform - the Enterprise Kubernetes distribution from SUSE - in a fully automated and orchestrated process.
This session will show how powerful is the integration and how quick and easy is the creation of a fully functional enterprise grade Kubernetes cluster that leverages all the capabilities of OpenStack.

Multi Tenant Networks for AI cloud with OpenStack

Networks for AI getting traction with workloads evolving for deep learning. Today’s AI networks needs high performance and needs all cloud related properties. On network side, the latest SDN revolution is centered on creating efficient virtualized data center networks using RDMA, VXLAN & EVPN. We will talk about AI network at the scale, performance, and cost advantages of using a modern controller-free Multi Tenant Networks for AI cloud with OpenStack" virtualized network solution built on 100 Gigabit Ethernet switches with hardware based VXLAN Routing. We will explore the ease of automating such a network in an OpenStack environment and take you through a real world use case of using OpenStack Network Node bridging between a bare metal cloud (EVPN) and a fully virtualized cloud environments (orchestrated by Neutron).

Why OpenStack upgrade is not a problem anymore for University of Edinburgh private cloud

One of the leading universities in the world, amongst top 6 in the UK is talking about the full OpenStack solutions the university recently adopted, including the Upgrade phase with zero-downtime upgrade.
Its cloud objective was to adopt a flexible infrastructure of compute, storage, network, services for the research IT infrastructure, without any service interruption. We know that the rapid release n to n+1 OpenStack upgrade every 6 months is a major challenge for many operators, but it is time to accept that it doesn’t require magic skills to solve this challenge.

Roberto di Lallo

Recycling keystone auth for kubernetes container platform through self service Application Credentials

This work aims to join the IaaS solution based on OpenStack with the container platform based on Kubernetes enabling OpenStack users to transparently log into Kubernetes. We present our user authentication deployment based on webhook to redirect the users’ requests towards the OpenStack Keystone, which is acting as auth-provider. We provide the details to obtain a solution usable from any cloud provider who wants to unify the authentication service starting from a IAAS infrastructure and expanding it into a container platform.

Session B

2:00 pm

Vincenzo Alonge

Datacenter evolution. Better availability and flexibility in a complex environment, with OpenStack architecture

A big leader in the financial industry is moving some business critical applications through a more flexible and resilient architecture. This section will show how challenging is the integration process within a really complex environment. Efficiency and Security are fundamental pillars of this initiative where all previous investments are protect.
DevOps means that people and processes must be involved for a better result.

Andrea Dell'Amico

Openstack at CNR-ISTI

The InfraScience Research Group is part of the Institute of Information Science and Technologies (ISTI), part of the Italian National Research Council (CNR).
The Institute is located in the CNR Research Area of Pisa.
More than 500 virtual machines (run by tenths of hypervisors and storage servers) serve a community of more than 4000 researchers from all Europe and beyond.
In this talk we present the complexities of a migration from Xen PV hypervisors to a hybrid cloud infrastructure backed by OpenStack and Ceph.
A totally manual lifecycle of VMs, apps and services has been completely automated using Ansible and Jenkins (CI/CD), reusing the already available hardware resources and with minimal services downtime.

Massimiliano Romano

AVNFM a powerful multi-purpose ETSI VNF Manager

The AVNFM, which has won the 2nd prize at NFV Innovation Hackathon organized by the EU Project SoftFIRE (, has the objective to introduce the Ansible tool inside NFV platforms allowing VNF deployment and the whole lifecycle management through the usage of Ansible Playbooks. The AVNFM goal is to fill the gap between customized self-developed ansible tools for NFV and the ETSI NFV specifications, allowing it to be integrated with any NFV-MANO platform compliant with ETSI specifications.

Stefano Stalio

OpenStack at INFN

INFN Corporate Cloud (INFN-CC) is a geographically distributed private cloud, based on OpenStack, hosted by three INFN data-centres. Its redundant architecture and its resiliency characteristics make of it the perfect environment for providing critical IT services and applications, while the fact that it is hosted in modern and large data-centres makes of INFN-CC the platform of choice for different scientific computing use cases.
A PaaS layer, developed within the EU funded project INDIGO-DataCloud (, allows for easier and federated access to computing and storage resources and automatic instantiation of scalable services and scientific applications.
We present the INFN-CC architecture, the challenges and advantages of our solution, and the first supported use cases.

Contrail Multicloud Security and Analytics for Containers

Juniper Contrail Enterprise Multicloud is a single platform that handles all overlay and underlay management; heterogeneous compute environments, including bare metal servers, virtual machines, containers, and networking devices; private and public clouds; networking and security orchestration policies, including micro-segmentation and advanced analytics. Contrail Enterprise Multicloud follows a pure software-defined approach that spans physical or virtualized public and private cloud infrastructures, providing Network as a Service (NaaS) for workloads running on physical, virtualized, or containerized form factors in any cloud environment. Contrail Enterprise Multicloud integrates with widely used orchestration tools such as OpenStack, Kubernetes, Mesos, OpenShift, and VMware cloud management systems, as well as popular DevOps tools like Ansible and Helm.

Simone Zanotti

TruStack – the OpenStack Easy, Powerful and Flexible, for everyone

Simone Zanotti
Who is Alchelabo, why they have created an OpenStack meta-distro, guiding principles, TruStack overview, technical deep dive and final live deploy demo.


Sala Busti

Session C


Luigi Annunziata

Openstack Cloud Security: Microsegmentation and software-defined security with Nuage VSS

Nuage Virtualized Security Services (VSS) is a software-defined security solution that is based on Nuage VSP platform to help address protection, detection, and operational challenges in cloud, data center, and branch environments.
Discover how you can secure east/west traffic inside data center, define and enforce whitelist policies for micro-segmentation, contextual visibility to application traffic for compliance monitoring, and define alerts with automated response in any Openstack Cloud environment.

Marco Paschetta

OpenStack for the Public Sdministration

CSI implements and manages IT Service for the public administration. The Nivola Program aims to offer to our stakeholders a cloud infrastructure based on Opensource solution such as Openstack.

Giovanni Merlino


Stack4Things is an open source umbrella project for deep IoT-Cloud integration empowering edge computing scenarios.
Our use cases focus around a number of Smart City-oriented verticals, under the #SmartME umbrella project, mostly featuring popular embedded and mobile systems as full-blown (far) edge nodes.
As part of the Stack4Things project, we developed a subsystem for IoT (bare metal) management, called IoTronic, that helps you in managing IoT device fleets at the edge without caring about their physical location, their network configuration, their underlying technology.
Recently IoTronic was promoted to the shortlist of projects the Edge Computing Group is agreeing upon to have the OpenStack ecosystem fully support the edge in the near future.

Roberto Reale

Blockchain and Openstack

My talk will examine the architecture of a blockchain application hosted on OpenStack, both in terms of requisites/challenges and of envisageable use cases in cloud governance models.

Session B

in Piccola Protomoteca

21 September




Session 1B
in Piccola Protomoteca


Speaker's bio.

Talks from the Call for speakers

1:00 pm

Terrazza Caffarelli



Session 2B
in Piccola Protomoteca


Speaker's bio.

Talks from the Call for speakers

4:00 pm

Coffee break
sala busti



Session 3B
in Piccola Protomoteca


Speaker's bio.

Talks from the Call for speakers


6:00 PM

Conclusions and end of sessions

Beyond OpenStack: GARR experience in designing and implementing a container Platform tailored on the Research community

In the world of public clouds, many companies are investing on containers, but this technology hasn't caught on with academia and research yet, despite the fact that some of its features are especially attractive for some categories of scientific applications. At GARR, we’re launching a container platform dedicated to the Italian Research community. We selected an architecture which allowed us to implement containers at the bare metal level. This approach caters for better performances and offers a direct access to the interfaces of specific devices, including GPUs (Graphics Processing Units), that cannot be shared with traditional virtualization techniques. Our platform highly relies on automation techniques, in order to to easily and quickly deploy the containers and reboot them at need, thus ensuring very high levels of fault tolerance.