Speakers
Description
INFN-CNAF is the national computing center of the INFN (National Institute for Nuclear Physics), dedicated to research and development in information technologies for subnuclear, nuclear, and astroparticle physics. CNAF hosts the largest INFN data center and operates a WLCG Tier-1 site.
For more than 15 years, tape data management at CNAF has been handled using the Grid Enabled Mass Storage System (GEMSS), an in-house solution. GEMSS middleware is based on a custom integration of the IBM Storage Scale file system with a tape backend powered by IBM Storage Protect, which provides Hierarchical Storage Management (HSM). This setup is further integrated with the Grid Storage Resource Manager (StoRM) through the StoRM WebDAV service and the StoRM Tape REST APIs. Today, this framework provides access to over 200 PB of data stored on tape and supports tens of scientific collaborations at CNAF.
After presenting a detailed overview of GEMSS architecture and operations, with particular emphasis on the stage-in and stage-out workflows, we describe how the system has evolved to support an arbitrary number of tape libraries and to enable dynamic data management on tape. We then discuss the enhancements introduced to efficiently exploit NVMe partitions in the tape buffer, together with the results achieved. Finally, we describe the implementation of a time-tracking mechanism for the main software components, aimed at improving monitoring and control of internal operations.