I stumbled across this particular project while browsing recent developments in supercomputer operations/observability at ORNL ( https://www.osti.gov/servlets/purl/1645803 ). Interestingly, per <https://ovis-hpcreadthedocs.readthedocs.io/en/latest/> and <https://github.com/ovis-hpc>, the whole system appears to be in active development and use across a number of HPC clusters in academia and in various labs. It has a particular focus on workload aware synchronised, approximately-zero-impact measurement gathering and machine/human decision operations and optimisation support. I was amazed that it hadn't appeared to have garnered more attention outside of academic/HPC circles ( plenty of formal papers/presentations however! - https://hmdsa.github.io/hmdsa/pages/docs.html )
1 comment
[ 0.20 ms ] story [ 1027 ms ] thread