This page shall give you an overview of the CITK infrastructure and what instances are connected to each other — including the designated purpose.
In the center of the CITK resides the catalog (*this). It provides interested users and developers with a) instructions on how to reproduce systems that are listed in the catalog, b) connects to multiple other platforms to receive, push or trigger information, build processes, and executable experiments — in simulation for instance — related to systems, included components and their associated artifacts, and c) semantically links all required entities relevant for reproduction. For detailed information please visit the Concept page.
The opensource.cit-ec.de server is the basis for recipe creation and maintenance. Moreover, it hosts various data sets and their meta data description which can be imported and linked here. For instance, please visit Open Data Amarsi Project and the corresponding catalog entry Amarsi Benchmark Framework
The Bielefeld University Library PUB service enables CITK users to import exisiting publication records (stored in PUB) and link them to systems, components, data sets ... Vice versa, the catalog provides a REST interface which is used by the library service to create backlinks to entries in the catalog. Lastly, CITK users may request a DOI for their system entry, the DOI management is also handled by the BU library.
The Continuous Integration Servers (A-C) serve multiple purposes. First of all, they provide means to frequently check the build status of single software components and entire systems (quality assurance). Secondly, information about the build statuses of diverse entities is imported and visualized in this catalog. Thirdly, experiment runs can be triggered from the catalog and watched be in realtime!Lastly, since we are using from source builds, as well as Linux containers, deploying a system should be manageable on almost every Linux derivative. However, for now we can provide full support for Ubuntu Linux 14.04 LTS only — also as the host system in a Linux container. This limitation almost only occurs because we need to manage and maintain basic operating system dependencies (*.deb packages) for every system and component in the CITK. If you are interested in deploying a CITK-enabled system on any other platform (even on Mac OSX) please feel free to contact us: citk-user [AT] lists DOT cit-ec DOT uni-bielefeld DOT de
The CITK promotes open data, software and science. The CITK, as a research project, is made feasible by using already existing open source building blocks, these (among many others) are: