Hardware infrastructure
The IT infrastructure is multi-site in order to guarantee the redundancy of the resources offered, and proposes the hosting of virtual machines under two OS (Debian Linux and FreeBSD) in order to limit the constraints imposed by a single OS. The number of virtual machines is growing rapidly (40+ machines under BSD and 23 under linux) currently. Project storage is quickly increasing (currently 400TB). The common calculation resources are co-administered by PADC and the DIO (Cluster and Grid).
The activities concern: the daily management of the archive, the storage and the calculation resources; OS update and verification test; the extension of the automated management of configurations through "puppet" facilitating disaster recovery plan.
A number of tools and dedicated server have been set up by PADC that can be used throughout the observatoire de Paris: Confluence wiki, Gitlab and SVN, ticket manager RT, NextCloud, service monitoring tool Nagios, monitoring and storage tool Munin.
Next Cloud has been used to share simulation results for SKA, Confluence for VESPA and CTA, and Gitlab is scaling up and integrating project management tools and continuous integration elements.
Actions are being planned for the implementation of a Docker type container infrastructure to minimize the time between the development of a prototype and the associated online service. Deployment on EOSC (European Open Science Cloud) is under assessment.