Hydra is down since Mon 4/22 9am until Fri 5/3
9am as we migrate the cluster to the Rocky 8.9 Linux
distribution and upgrading the job scheduler (Grid Engine) to v8.8.1 &
the OS management tool (BCM) to v10.0.
All running jobs have been terminated and all queued jobs deleted.
During that time all logins will remain disabled, and there will be no
access to any files stored on Hydra.
We will reopen Hydra as soon as this migration and its validation are completed
and will announce its availability.
We are also upgrading various packages to their most recent versions and
reorganizing & renaming many of modules. Old versions of some software
packages are no longer supported.
We are adding 15 new compute nodes (2 nodes with 192 CPUs and 1.5TB of
memory, 12 nodes with 128 CPUs and 1.0TB of memory, and 1 node with 4 GPUs -
NVIDIA L40S, 48GB).
The composition of the cluster will grow to more than 5900 CPUs, over 44TB
of aggregate memory, distributed over 78 nodes, with 3 nodes with GPUs (one
quad, two dual, or 8 GPUs).
We are also consolidating the public disk space (fewer volumes/filesets) and
increasinf user quotas.
We are in the process of updating the
'2024
Cluster Upgrade to Hydra-7' page that describes what has changed
including module reorganization/naming and new versions.
The documentation at
the Wiki
will be updated as soon as we can - tagging pages with 'updated for
hydra-7'.
We plan to offer a workshop or 'office hours' to help users to transition.
As always we are striving to make this transition as smooth as possible,
while leveraging the opportunities and challenges of using a new version of
the OS.
|