With the start of the summer term bringing with it MSc and undergraduate projects, some improvements are in the pipeline:
tape backups: increasing concern over the possibility of ransomware attacks, loss or theft of research data and the deteriorating geopolitical situation in Europe has resulted on renewed focus on off-line backups. These have the advantage over automated mirroring of storage servers of an "air break"; a tape that is not loaded into a tape drive simply cannot be overwritten or tampered with since there is nothing but air between IT systems and the tape! More backup schedules have been set up on the automated tape library system to create additional backups of sensitive and/or irreplaceable data and we are now awaiting delivery of more tape cartridges.
development of a new Hadoop/Spark platform: first mentioned in the March update, this work has been completed with the introduction of aphrodite, a single server implementation of Apache Hadoop and Spark designed to meet the spacifications for a machine learning/AI course that started this term. Eventually, the big Bazooka Hadoop cluster will be rebuilt in the same way, based on modern Python 3 software with enhanced remote access facilities for off-campus users.