Nvidia

See merge request !165

Former-commit-id: 1d8ec82a

Former-commit-id: 5cc65913

Former-commit-id: 7ec5dac7

Former-commit-id: b8afca88

Upgrade

See merge request !164

Former-commit-id: 8e2c23ba

Former-commit-id: d518678c

changes to the config_repos role so hopefully we can upgrade cleanly without reinserting the base repos


Former-commit-id: 2c9accd0

M3 has decided to change where we store some parts of slurm, in particular prolo…

See merge request !163

Former-commit-id: 5175eccb

Slurm epilog

See merge request !161

Former-commit-id: 064ba596

Former-commit-id: fa156ea5

Former-commit-id: 1cd81921

M3 has decided to change where we store some parts of slurm, in particular prolog and epilog will go into /opt/slurm/etc


Former-commit-id: e6eb57fb

update the unit file for slurmdbd to respect the directory we want the pid to go to

See merge request !159

Former-commit-id: 50906272

Adding option to disable KMEM in cgroup

See merge request !160

Former-commit-id: 8497a988

Former-commit-id: aac097c0

Former-commit-id: ce047d0c

allow use of a ldap readonly server for failover (but don't mandate it)

See merge request !158

Former-commit-id: 0db598c8

Former-commit-id: 59d9297a

set the killmode so restarting slurmd doesn't cause the users jobs to be killed

See merge request !157

Former-commit-id: 69cbefd0

Former-commit-id: e467abe0

Remove -a to make the check_fs_used work

See merge request !156

Former-commit-id: c59547d9

Refer to https://github.com/mej/nhc


Former-commit-id: 87dd7d06

Former-commit-id: fd0db365

Do not enabled slurmctld and slurmdbd in systemd

See merge request !155

Former-commit-id: 48b1f2bb

Make slurmctld and slurmdbd explicitily disabled on startup (rather than just leaving it to whatever it is currently set to)

Former-commit-id: adb46795

Former-commit-id: b5466172

fix an error detecting the ppolicy and enable the auxposixgroup schema

See merge request !154

Former-commit-id: 0da2112b

Former-commit-id: 7519d863

Fix nat server

See merge request !153

Former-commit-id: 855639ba

Former-commit-id: 2a302213

Former-commit-id: 80ddd671

Former-commit-id: 7a126398

Merge branch 'master' of gitlab.erc.monash.edu.au:hpc-team/ansible_cluster_in_a_box into fix_nat_server

Fix nat server to point iptables to use correct public interface name


Former-commit-id: fc87ec4f

- template the public interface so NAT Masquerade command goes to right interface
- public inteface name autogenerated by looking at route to 8.8.8.8
- ACCEPT command for private interface modified to use network CIDR. THis is hardcoded to M3 if not defined, so no impact on cluster_in_a_box roles. (Monarch and others will need to pass it as a parameter)
- Not sure if ACCEPT on private interface is used, as compute nodes point to mlx0 interface for routing


Former-commit-id: 93dc59ba

Former-commit-id: aa5a5dec

Fixed xorg.conf.j2 template

See merge request !150

Former-commit-id: 2628cd87

Former-commit-id: eab73154

added gpu role to generate combinations of xorg.conf

See merge request !149

Former-commit-id: 4d4f03d6

Former-commit-id: 5848d88f

added check_nvidia_uvm.nhc

See merge request !148

Former-commit-id: 864bc2d3