Merge branch 'main' into filesystems-docs

Author: sjpb

ansible/roles/cuda/defaults/main.yml

@@ -1,7 +1,7 @@
 cuda_repo_url: "https://developer.download.nvidia.com/compute/cuda/repos/rhel{{ ansible_distribution_major_version }}/{{ ansible_architecture }}/cuda-rhel{{ ansible_distribution_major_version }}.repo"
-cuda_nvidia_driver_stream: '575-open'
-cuda_nvidia_driver_pkg: "nvidia-open-3:575.57.08-1.el{{ ansible_distribution_major_version }}"
-cuda_package_version: '12.9.1-1'
+cuda_nvidia_driver_stream: '580-open'
+cuda_nvidia_driver_pkg: "nvidia-open-3:580.65.06-1.el{{ ansible_distribution_major_version }}"
+cuda_package_version: '13.0.0-1'
 cuda_version_short: "{{ (cuda_package_version | split('.'))[0:2] | join('.') }}" # major.minor
 cuda_packages:
   - "cuda-toolkit-{{ cuda_package_version }}"

ansible/roles/hpctests/library/plot_nxnlatbw.py

@@ -129,7 +129,11 @@ def run_module():
             if len(vals) != 4:
                 print('warning: skipping line %i (%i values)' % (ln, len(vals)))
                 continue
-            rankA, rankB, lat, bw = int(vals[0]), int(vals[1]), float(vals[2]), float(vals[3])
+            try:
+                rankA, rankB, lat, bw = int(vals[0]), int(vals[1]), float(vals[2]), float(vals[3])
+            except ValueError:
+                print('warning: skipping line %i (%s) - parse failure' % (ln, line))
+                continue
             latencies[rankA, rankB] = lat
             bandwidths[rankA, rankB] = bw
 

environments/.caas/ansible.cfg

@@ -54,7 +54,8 @@ openhpc_config_extra: {}
 # additional default slurm.conf parameters for the appliance:
 openhpc_config_default:
   TaskPlugin: task/cgroup,task/affinity
-  ReturnToService: 2 # workaround for templating bug TODO: Remove once on stackhpc.openhpc v1.2.0
+  ProctrackType: proctrack/cgroup
+  JobAcctGatherType: jobacct_gather/cgroup
 
 # additional default slurm.conf parameters when "rebuild" enabled:
 openhpc_config_rebuild: