HPCasCode merge requestshttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests2021-06-16T14:33:29+10:00https://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/472replaced file with copy2021-06-16T14:33:29+10:00Jafar Liereplaced file with copyTwo changes here:
1. Replace file with copy (to initiate the file transfer)
2. Remove the leading files/ from slurmctld.prolog path, the prolog file is already in the files/ directory and ansible searches this directoryTwo changes here:
1. Replace file with copy (to initiate the file transfer)
2. Remove the leading files/ from slurmctld.prolog path, the prolog file is already in the files/ directory and ansible searches this directoryAndreas HamacherAndreas Hamacherhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/470fixing hostgroups2021-06-07T11:25:06+10:00Andreas Hamacherfixing hostgroupsSimon MichnowiczSimon Michnowiczhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/471well it does not get more ugly than his :( I don't have a reliable way of...2021-06-07T11:24:49+10:00Andreas Hamacherwell it does not get more ugly than his :( I don't have a reliable way of...well it does not get more ugly than his :( I don't have a reliable way of getting this working so I am leaving residuals of debbuging and workaround codewell it does not get more ugly than his :( I don't have a reliable way of getting this working so I am leaving residuals of debbuging and workaround codehttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/462working on local cache only, so having no internet does not cause a timeout2021-06-07T11:18:06+10:00Andreas Hamacherworking on local cache only, so having no internet does not cause a timeouthttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/465making nat work on ubuntu20 hopefully2021-06-04T11:14:31+10:00Andreas Hamachermaking nat work on ubuntu20 hopefullythere will be a second mr for clusterbuild soon which will allow you to test thisthere will be a second mr for clusterbuild soon which will allow you to test thisTrung NguyenTrung Nguyenhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/463changing repository pathing to suit a specific centos version2021-06-04T11:13:52+10:00Andreas Hamacherchanging repository pathing to suit a specific centos versionTrung NguyenTrung Nguyenhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/469rootless docker role. Ubuntu only for now2021-06-04T11:05:04+10:00Andreas Hamacherrootless docker role. Ubuntu only for nowtested on m3-ubuntutestlogin
no RHEL support yet. although Centos 7 would be supported. I think this should be discussed widely before thinking about a roleouttested on m3-ubuntutestlogin
no RHEL support yet. although Centos 7 would be supported. I think this should be discussed widely before thinking about a roleoutTrung NguyenTrung Nguyenhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/468Update roles/slurm-common/files/slurmctld.prolog2021-06-04T07:49:10+10:00Chris HinesUpdate roles/slurm-common/files/slurmctld.prologhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/467installing missing library2021-06-03T10:27:19+10:00Andreas Hamacherinstalling missing libraryhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/435Ubuntu related changes2021-06-02T09:42:00+10:00Andreas HamacherUbuntu related changeshttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/466adding a slurmctld.prolog2021-06-01T10:32:09+10:00Andreas Hamacheradding a slurmctld.prologhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/464Fix listdeps2021-05-27T12:49:40+10:00Luhan ChengFix listdepsFixing two issues
* It didn't work if only module name is passed to the script
* It was only searching for default versions, I updated the script so it searches for all versions
A sample of `./listdeps openmpi` output
```
abinit/8.8.3...Fixing two issues
* It didn't work if only module name is passed to the script
* It was only searching for default versions, I updated the script so it searches for all versions
A sample of `./listdeps openmpi` output
```
abinit/8.8.3
amber/18-multi-gpus
amber/18-parallel
amber/18-parallel-pmemd.gem
chimera/1.14
cnvnator/0.4.1
cp2k/5.1.0
cp2k/6.1.0
cpmd/3.17.1
cpmd/4.3
dense3dcrf/20160527
exciting/nitrogen
gamess/16srs1
gamess/16srs1-v2
gamess/2018r3
geant4/10.6
gromacs/2016.3-openmpi-cuda8.0
gromacs/2016.4-openmpi-cuda8.0
gromacs/2016.5-openmpi-cuda8.0-plumed
gromacs/2018.4-openmpi-cuda8.0
gromacs/2018.7-openmpi-cpu-only
gromacs/2018.7-openmpi-cuda-plumed
gromacs/2018-openmpi-cuda8.0
gromacs/2018-openmpi-cuda8.0-NVML
gromacs/2019.4-openmpi-cuda10.0
gromacs/2020.1-openmpi-cuda10.1
gromacs/5.1.4
hypre/2.11.2
hypre/2.15.0
iqtree/1.6.2
lammps/20180510
lammps/20181212
lammps/20200303
lammps/20200303-python3
mpifileutils/20200701
namd/2.13-multicore-CUDA
....
```Andreas HamacherAndreas Hamacherhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/461Ubuntu20cicd2021-05-24T10:42:30+10:00Andreas HamacherUbuntu20cicdThe purpose of this branch is to have a working CICD pipeline with a ubuntu20 VM image and HaC ontopThe purpose of this branch is to have a working CICD pipeline with a ubuntu20 VM image and HaC ontophttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/460Upcoming maintenance2021-05-21T12:34:51+10:00Andreas HamacherUpcoming maintenanceTrung NguyenTrung Nguyenhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/459Slurm20cicd2021-05-19T13:06:26+10:00Andreas HamacherSlurm20cicdthis branch is introducing some breaking slurm20 changes which are compatile with monarch at this point in time but not M3
The other purpose is to bring back the CICD pipeline to a green status to allow for ubuntu related commits to be ...this branch is introducing some breaking slurm20 changes which are compatile with monarch at this point in time but not M3
The other purpose is to bring back the CICD pipeline to a green status to allow for ubuntu related commits to be properly checked.
I had to revert some dummy ldap passwords to redhat
The changes in `roles/slurm-start/tasks/main.yml`
are very spagetti.
The core problem is that we have one role for slurm controllers and clients and also that we want the CICD pipeline to start everything automagically and he live systems to be only started manuallyhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/458Lmod changes2021-05-13T16:02:29+10:00Andreas HamacherLmod changessmall changes I made to get lmod working while templating out Redhat lmod files as well
@jlie0003 I ended up having to create a new branch and cherry picking your first commit.
I stuffed up the other branch.
Please verify the below. ...small changes I made to get lmod working while templating out Redhat lmod files as well
@jlie0003 I ended up having to create a new branch and cherry picking your first commit.
I stuffed up the other branch.
Please verify the below.
And yes this is the extend I go to to avoid code duplication. Maybe I am overtrained to avoid it...https://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/456insert a random sleep for these cron jobs so all nodes are not executing at t...2021-05-11T13:32:23+10:00Chris Hinesinsert a random sleep for these cron jobs so all nodes are not executing at the same timehttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/452IP change of mysql-backups2021-05-07T09:54:28+10:00Andreas HamacherIP change of mysql-backupsmore to come ( e.g. hpcuser ) but lso clusterbuildmore to come ( e.g. hpcuser ) but lso clusterbuildPhilip Chanphilip.chan@monash.eduPhilip Chanphilip.chan@monash.eduhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/455fixes for lint. no changes to logic2021-05-06T15:54:50+10:00Andreas Hamacherfixes for lint. no changes to logicAndreas HamacherAndreas Hamacherhttps://gitlab.erc.monash.edu.au/hpc-team/HPCasCode/-/merge_requests/454Remove probe for GRES/CPU on gpu nodes, and remove gres.conf & slurm.conf, go...2021-05-06T13:40:51+10:00Damien LeongRemove probe for GRES/CPU on gpu nodes, and remove gres.conf & slurm.conf, going configlessAndreas HamacherAndreas Hamacher