e-infrastructural needs to support informatics

Editor's Notes

• #6: Hosted through the Centre for Environmental Data Analysis(CEDA) at STFC the JASMIN service[13] is intended to be a multi user and multi use-case facility that is able to support a wide range of user communities. The system is built in a modular fashion with a hardware layer that is deliberately designed to support multiple different distributed computing paradigms, from HPC type workloads to cloud hosting. Utilising commercial off-the-shelf private cloud software, which supports API level access, the cloud service operated in two modes: firstly a highly controlled Platform as a Service system in the managed cloud, and secondly a more flexible community or user controlled Infrastructure as a Service system, the unmanaged cloud. Users may opt to run in the managed mode, where VMs are standardised to JASMIN approved templates and configured by Puppet[14] scripts that apply centralised access controls, or in the unmanaged mode which we use for the EOS Cloud. Here each user can have full root access rights on their VM once it is deployed and handed over to them, and we as clients of JASMIN can make use of the full vCloud Director API[15] to assign resources to VMs according to our own policy. A further advantage of working on JASMIN is that, as a community cloud, the service already hosts, and synchronises data from relevant providers. These include large reference datasets which our target user community finds useful. Users can rely on this resource and only need to concern themselves with the transfer of their own data into and out of the cloud.
• #9: Bio-Linux[1] is a major success with widespread adoption across multiple communities. It provides a method by which personal bioinformatics workstations can be set up and maintained by novice users, removing the need for individuals to compile and install packages themselves. The project has an active user and developer community and engages with other open-source community efforts like Debian Med[2], Galaxy[3], Open Bioinformatics Foundation[4], etc. Workstations running Bio-Linux connect to a package repository where hundreds of tools have been built into Debian (.deb) packages and are tested and continually updated by the community of maintainers. Installation or removal of even complex tools is trivial for the end user, and updates are automatic. Bio-Linux has traditionally been installed onto bare-metal, possibly on a dual-boot workstation, or run directly from a USB stick for training and demo purposes. In recent times, use on local Virtual Machines(VM) and Cloud systems has increased, with a community CloudBioLinux[5] effort building and releasing machine images on the Amazon EC2[6] public cloud. Also, partly through this current project, support for local VM installations (on VirtualBox[7], Parallels[8], VMWare[9]) has been made a core part of each Bio-Linux release. Users can obtain an Open Virtualization Archive (OVA)[10] file which loads directly into these products to instantly provide a functional desktop system. It is this same pre-packaged virtualised workstation that users now access on the EOS Cloud.
• #10: Docker[11] allows a uniform platform-independent application to be built and encapsulated in a format know as a container. As the research community, and people in general, make use of a large number of different operating systems the complexity of installing applications on new systems and can be high. Sometimes there may be incompatibilities between the libraries and helper tools needed by an application and those on the target platform. The Docker approach aims to get round this by providing single application hosting platform irrespective of hosting system/operating system etc. We have recently seen within the bioinformatics community an increase in sharing of applications and tools in this manner. One problem faced when using the Docker hosting environment is that the applications that are installed and running within it look and feel like remotely hosted applications, they do not behave as if they have been installed on the users machine. Therefore a tool that bridges this usability gap would greatly increase the acceptability of tools and services that are installed and distributed using Docker.
• #14: Hyun Soon Gweon and Katja Lehmann, two bioinformaticians as CEH, are familiar with using Bio-Linux on the departmental server and personal VirtualBox VMs. They are trialling the system as part of their regular work and to provide a shared computing environment with project collaborators in microbial community profiling. Ranny van Aerle at University of Exeter is unable to run a Bio-Linux workstation at his home institute due to local IT policies and budget constraints. He needs a basic workstation plus occasionally some extra computing power. While he is not at present being charged for use of the pilot system, he represents exactly the type of user for whom a paid subscription to the service would make economic sense. Since being given access to a Bio-Linux instance on EOS Cloud in October 2014 he has used it to assemble bacterial genomes, annotate large metagenomics datasets using Blast and Diamond, visualise Blast/Diamond results in MEGAN and generally to try out new software and scripts.
• #16: The Software Sustainability Institute can help with: software reviews and refactoring, collaborations to develop your project, guidance and best practice on software development, project management, community building, publicity and more… Drawing on pool of specialists to drive the continued improvement and impact of research software developed by and for researchers Providing services for research software users and developers Developing research community interactions and capacity Promoting research software best practice and capability
• #17: KEY POINT: Get across what the Institute does This is a conceptual breakdown to make the Institute’s activities easier to manage. In reality the Institute staff work across themes, contributing to many activities, as software sustainability cannot be addressed just through individual activities working alone. Specific activities: Software Carpentry courses teach researchers key computational skills Helping to spin up ELIXIR-UK training activities DIRAC Drivers License to get best use out of HPC resources Advising Wellcome Trust on software policy Guides on wide range of subjects We understand the linkage between all of these requirements, issues and activities because we’ve investigated the research communities reliance on software in depth.