Possible contributions of UPV to WP5 (NDIGO- DataCloud) Germán Moltó UPV RIA
UPV as a Partner of INDIGO Universitat Politècnica de València (UPV) Institute for Molecular Imaging Technologies (I3M) Grid and High Performance Computing Group (GRyCAP) European Projects EGEE, EELA, VENUS-C, EGI-* INDIGO-DataCloud RIA 20+ researchers in areas such as Cloud Computing, Biomedicine or Scientific Computing
IaaS Managing Services IM CloudVAMP EC3 VMCA VMRC CLUES A Vertical Elasticity IaaS Memory Overcommitment Manager for On-premise Clouds A Virtual Machine Consolidation Agent An energy management system for Clusters and Cloud infrastructures. Automatic configuration and recontextualization service. Creation of elastic virtual clusters on top of both public and on-premise IaaS providers A rich metadata repository of Virtual Machine Images. Open-Source developments available at
VMRC A Virtual Machine Image Catalog Manager To enable users to catalog their VMIs so that they can later be searched, retrieved and used. Metadata: Operating System name, flavour and version, installed applications, etc. URI naming convention to abstract from cloud provider: one://server:port/image-id ost://server:port/ami-id aws://region/ami-id gce://region/image-name … Federation Index similar VMIs in different on-premises Clouds INDIGO-DataCloud RIA
Infrastructure Manager Open source platform to deploy on demand customizable virtual computing infrastructures. With the precise software configuration required. Allow to deploy any kind of complex infrastructure. No need of pre-configured VMIs. Enable re-using of VMIs. The same complex infrastructure can be deployed both on-premise and in a public Cloud. Enables platform-agnostic or hybrid deployments INDIGO-DataCloud RIA Modular design enabling: Currently supports: OpenNebula, OpenStack, EC2, GCE, Azure, OCCI (FedCloud), FogBow, Docker, LibVirt. Service with two APIs: XML-RPC and REST.
Example: Hadoop Cluster network privada network publica (outbound = 'yes’) system front ( cpu.arch='x86_64' and cpu.count>=2 and memory.size>=2048m and net_interface.0.connection = 'publica' and net_interface.1.connection = 'privada' and net_interface.1.dns_name = 'hadoopmaster' and disk.0.os.name='linux' and disk.0.os.flavour='ubuntu' and disk.0.os.version='12.04' and disk.0.applications contains (name='ansible.modules.micafer.hadoop') ) system wn ( cpu.arch='x86_64' and memory.size>=2048m and net_interface.0.connection='privada' and disk.0.os.name='linux' and disk.0.os.flavour='ubuntu') configure front - roles: - { role: 'micafer.hadoop', hadoop_master: 'hadoopmaster', hadoop_type_of_node: 'master' ) configure wn - roles: - { role: 'micafer.hadoop', hadoop_master: 'hadoopmaster' ) deploy front 1 deploy wn 5
Elastic Cloud Computing Cluster (EC3) EC3 is an open-source tool to deploy self-managed virtual elastic clusters across hybrid Cloud infrastructures Elasticity controlled by CLUES ( INDIGO-DataCloud RIA Advantages No upfront investment. Usage-dependent cost. Customizable. Cost-aware Automatic elasticity. Hybrid scenarios (on-premises + public Cloud
Possible Contributions to WP5 T5.1/T5.3 Elasticity/Scalability Management T5.3 High Level Application Deployment E.g. Hadoop Clusters INDIGO-DataCloud RIA
Contacts Germán Moltó, Ignacio Blanquer, Miguel Caballer Instituto de Instrumentación para Imagen Molecular Universitat Politècnica de València Camino de Vera s/n 46022, Valencia SPAIN INDIGO-DataCloud RIA