Performance enhancement results – Hybrid Cloud

Copyright © 2014 Open Geospatial Consortium. 33 Figure 13 - Average response time according to the number of simultaneously served requests in a Hybrid Cloud deployment source [2] The experiments show that a significant performance enhancement can be reached by the Hybrid Cloud approach.

5.3 Performance enhancements of DinSAR processing using a Hybrid Cloud

The principle of the Differential SAR Interferometry DinSAR is illustrated Figure 14, and can require large storage and compute-intensive scientific processing, as it is the case for the Small Baseline Subset SBAS technique involved in this testbed deployment. In this use case, we will exploit 64 differential SAR scenes for the generation of time series showing ground displacements over a decade in geological sensitive areas. This is part of an ongoing effort from ESA, CNR-IREA and Terradue partners to deliver updated and comprehensive time series over geologic phenomena, supporting scientific advising towards decision makers. This SBAS Cloud deployment testbed is focusing on the Naples area in Italy, making use of about 80 GB of ENVISAT ASAR scenes, generating 300 GB of intermediary and output products, and representing a time series of 9 years 2002-2010 that will show the sensitivity to ground displacements in the region. 34 Copyright © 2014 Open Geospatial Consortium. Figure 14 - The SAR Interferometry technique involved with the SBAS Cloud deployment The Hybrid Cloud deployed by Terradue Srl for this project makes use of the private Cloud deployed at Terradue’s facilities, along with public Cloud resources provisioned at the Amazon Cloud Compute infrastructure in Dublin, using their AWS API, and provisioned at the Interoute Cloud Storage Infrastructure in London, using Terradue’s Virtual Archive Cloud Service a SAR data archive deployed as the contribution of the European Space Agency to the GEO Supersites Exploitation Platform, http:eo-virtual- archive4.esa.int .

5.3.1 Scope of work

This work demonstrates performance enhancements for a Cloud deployment of the SBAS Small Baseline Subset processing application with the use of WPS and OpenSearch OGC Web Services for the production of On-demand Ground Deformation Maps. The activity was conducted in OGC Testbed 10 by Terradue, in collaboration with CNR- IREA SBAS processing chain and with grant program from Amazon Web Services Education Government Solutions. The deployment is exploiting resources from Terradue Cloud Controller, AWS public Cloud, Interoute public Cloud, and the GEO Supersites Virtual Archive Cloud Storage of massive ESA SAR data.

5.3.1.1 Scientific data processing leveraging Cloud Computing

The European Space Agency is a leading provider of Earth Observation scientific data. As a European contribution to GEO http:www.earthobservations.org , ESA is operating the SSEP Virtual Archive Service.