Authors: Philip C. Church; Andrzej M. Goscinski
Addresses: Faculty of Science and Technology, School of Information Technology, Deakin University, Geelong VIC 3127, Australia ' Faculty of Science and Technology, School of Information Technology, Deakin University, Geelong VIC 3127, Australia
Abstract: Through the use of cloud platforms, software packages have been centralised and made available as services. Utilising these software services require that data is retrieved from public and private data servers (called biological databases). These data servers are: difficult to discover, lack standardised access, and are time consuming to manage. In response, we propose a new approach in which data servers and files are abstracted in a similar manner to software services. This new approach uses meta-data to represent data servers and files as services. Data servers and files are discoverable and managed through a broker. Each data server service encapsulates the protocols required to access, and transfer data. When integrated into a cloud platform, this approach allows for users to take advantage of cloud, cluster, and biological database resources in a uniform manner, via web interfaces.
Keywords: data as a service; DaaS; software as a service; SaaS; bioinformatics; cloud computing; metadata; clusters; web interfaces; data access; data transfer.
International Journal of Cloud Computing, 2016 Vol.5 No.4, pp.264 - 282
Received: 31 Oct 2015
Accepted: 23 Nov 2015
Published online: 09 Dec 2016 *