This past month we skipped the TWG due to the lack of sufficient agenda items. The development team continued work on migration to our new CI system, iRODS 4.2 and the soon to be released iRODS 4.1.6. The Consortium has also traveled to the University of Utrecht for two days of advanced iRODS training.
We have completed migration of the GSI authentication plugin to the new jenkins system. This leaves the Kerberos plugin and the final clean up of the HPSS plugin. The next steps for CI will be inclusion of automated Federation testing, and testing of the Run In Place deployment.
Progress continues for iRODS 4.2. We are close to pushing the first implementation of the fully refactored python configuration suite and have implemented a trial repository for the Debian based package managers. We ran into a small roadblock for the rpm based systems, but are working with the developer of our chosen tool: PRM. Once this step is complete, we will build packages for all of our external dependencies.
We have collected 23 issues for the 4.1.6 release, which was driven primarily by an issue with the development libraries. We have now further partitioned the iRODS client library into a specific library for plugins as well as a client core library. This partitioning is another step closer to a dynamically linked iRODS build where we may maintain functionality in individually versioned dynamic shared objects.
Advanced Training in Utrecht
Dan and I traveled to the University of Utrecht where we were hosted by the Research Services and Data Management Group. We presented two days of advanced iRODS training, focusing on writing rules, microservice plugins, resource composition and data grid management. We had a lot of fun working with everybody, debugging broken demos, and discussing everyone’s individual use cases.