Skip to content

Commit e342812

Browse files
committed
Fixups for slurm recompile script
1 parent 5ac818d commit e342812

File tree

1 file changed

+6
-3
lines changed

1 file changed

+6
-3
lines changed

ansible/fatimage.yml

Lines changed: 6 additions & 3 deletions
Original file line numberDiff line numberDiff line change
@@ -254,19 +254,22 @@
254254
hosts: cuda
255255
become: yes
256256
tasks:
257-
- name: Get facts about Cuda installation
257+
- name: Get facts about CUDA installation
258258
import_role: cuda
259259
tasks_from: facts.yml
260260

261261
- name: Recompile and install slurm packages
262262
shell: |
263+
#!/bin/bash
263264
dnf download --source slurm-slurmd-ohpc
264-
rpm -i slurm-ohpc-23.11.10-320.ohpc.3.1.src.rpm
265+
rpm -i slurm-ohpc-*.src.rpm
265266
dnf install -y @'Development Tools'
266267
cd /root/rpmbuild/SPECS
267-
dnf builddep -y slurm.spec
268+
dnf builddep -y slurm.spec
268269
rpmbuild -bb -D "_with_nvml --with-nvml=/usr/local/cuda-{{ cuda_facts_version_short }}/targets/x86_64-linux/"
269270
dnf reinstall /root/rpmbuild/RPMS/x86_64/*.rpm
271+
# Workaround path issue: https://groups.google.com/g/slurm-users/c/cvGb4JnK8BY
272+
ln -s /lib64/libnvidia-ml.so.1 /lib64/libnvidia-ml.so
270273
271274
- name: Run post.yml hook
272275
vars:

0 commit comments

Comments
 (0)