…k` ,set storage port mode based on `cluster_storage_vnic_type`

* bump Packer source image to RL9.4

* downgrade OFED to LTS to get stable download url

* bump OOD role, now ondemand dnf package installed will be latest

* Revert Packer source image to RL9.3 to avoid hanging after post-update reboot"

This reverts commit 851c494.

* bump OFED to get RL9.4-supported version

* bump leafcloud packer vm to 8GB RAM

* DEBUG: disable (working) OFED build

* Revert "DEBUG: disable (working) OFED build"

This reverts commit 45a48c3.

* DEBUG: output builder hostname

* Revert "DEBUG: output builder hostname"

This reverts commit 3f95f8e.

* fix build workflow concurrency

* DEBUG: disable updates

* Revert "DEBUG: disable updates"

This reverts commit 3581a35.

* bump packer build volume size for non-ofed to avoid RL8 build running out of root space

* try to prevent stackhpc env image build connection drops

* bump packer source image to fixed RL9.4 image

* run test CI workflow on RL8 image if PR labeled with 'RL8'

* bump CI images

* bump openhpc role to fix munge checks on key path

* support ceph quincy for RL9

* bump CI image

* disable grafana repos by default

* bump CI image

* add squid role + config

* make systemd proxy config idempotent (and clearer)

* fix defaults for squid/builder

* bump CI image

* fix squid cache dir definition error

* Fix squid README nits

Co-authored-by: Scott Davidson <[email protected]>

* Fix squid README nits

Co-authored-by: Scott Davidson <[email protected]>

* Fix squid README nits

Co-authored-by: Scott Davidson <[email protected]>

* Fix squid README nits

Co-authored-by: Scott Davidson <[email protected]>

* make proxy default to squid port

---------

Co-authored-by: Scott Davidson <[email protected]>

* upgrade ssh from SIG/security to fix CVE-2024-6387

* refactor ssh update from sig/security to work on existing fatimage

* revert to base ssh repos

* fix Rocky-9-GenericCloud-Base-9.4-20240523.0.x86_64.qcow2 /etc permissions

* bump CI image

* TuneD configuration role

* bump CI image

* Update README.md

* Update environments/common/layouts/everything

Co-authored-by: Steve Brasier <[email protected]>

* Update ansible/roles/tuned/README.md

Co-authored-by: Steve Brasier <[email protected]>

* Update ansible/roles/tuned/README.md

Co-authored-by: Steve Brasier <[email protected]>

* Update ansible/roles/tuned/README.md

Co-authored-by: Steve Brasier <[email protected]>

* Update README.md

---------

Co-authored-by: Bertie <[email protected]>
Co-authored-by: Steve Brasier <[email protected]>

* use run_number as a shorter ID for CI

* slurmci group name warning

* Revert "slurmci group name warning"  - underscores not valid linux hostname and stripped in host, leading to slurmdbd config failure

This reverts commit 61dfad6.

---------

Co-authored-by: Bertie <[email protected]>

…fy image_id for some compute groups (#412)

* Add ansible-init role to requirements.yml

* Add ansible-init to groups and plays

* Configure cluster_infra ansible-init metadata

* Only run site.yml once ansible-init has completed

* Wait for ansible init to finish before running bootstrap

* revert to using cluster_infra metadata defaults

* update image

* revert sausage bastion changes

* set ansible_init_wait as common var

* use run_number as a shorter ID for CI

* install ood apps in fatimage

* add ood jupyter install to fatimage

* bump image

* jupyter_compute ood into fatimage

* bump fatimage for jupyter_compute ood

* Update stackhpc.yml

* duplicate tuned inventory group name

* Fix invalid group name for slurmci

* Update stackhpc.yml

undo groupname changes

* slurmci group name warning

* rm ood changes

* bump fatimage

* change azimuth collection in bootstrap

* update azimuth image utils version

* update requirements

* Update bastion.yml

* Use azimuth image utils collection for ansible-init

* bump fatimage

---------

Co-authored-by: bertie <[email protected]>
Co-authored-by: bertiethorpe <[email protected]>

* Check major version for RL8 package installs

* Gather facts on ofed role

* Support kernel checks with mismatching version length

4.18.0-553.16.1.el8_9.x86_64
4.18.0-553.el8_9.x86_64
These would fail with the error:

'<' not supported between instances of 'str' and 'int'.

as the community.general.version_sort was trying to compare the `el8_9` of the latter with the `16` of the former.

Strip the last two chunks so we just compare numbers.

* Move to LTS version now RL9.4 is supported

* Fail when any inventory source cannot be parsed

* Always reboot after selinux and package updates

* Cleat facts before OFED so install will match newest kernel

* Clear facts after reboot so OFED install will match newest kernel

* fail caas and stackhpc if any inventory can't be read

* make reboot conditional on package or SELinux changes again

* include OFED in both RL8 and RL9 builds

* always run CI tests on RL8 and RL9

* allow concurrent RL8/RL9 CI tests

* mark pending reboot check as not a change

* fix workflow matrix definitions

* bump CI images - now both OFED

* use reboot hint for checking reboot required

---------

Co-authored-by: Steve Brasier <[email protected]>

extract fatimage logs and process ansible timings

* configure SMS deployment

* add slurm-app-ci user

* bastion config

* allow overriding CI_CLOUD for PRs using label

* choose cloud for fatimage workflow_dispatch

* packer build qcows

* bump fatimage

* update packer common vars

* Update fatimage.yml

* Update main.tf images

* Update fatimage.yml

* revert fatimage.yml changes

* Update fatimage.yml

* Update fatimage.yml

* hcl packer var put string in quotes

---------

Co-authored-by: Steve Brasier <[email protected]>

Before we can move to v3.0.0 we need to use the new floating ip
resource, now the compute one has been removed.
To unblock this move, we add a tempory pin to use v2.1.0
for a bit.

* add description of image description to build

* fix image summary dir logic

* bump CI image

* Update openstack.pkr.hcl

* new image build workflow

* dynamically set packer vars from fatimage workflow

* remove openstack. prefix from image name

* echo image name

* make image_name var in packer config

* new changes

* fix merge changes

* temp workflow changes

* test nightly build

* change back fatimage workflow

* rename images built

* add update to builder group

* add update to fatimage build groups

* fatimage.yml fix

* move output image_name declaration into build blocks

* delete outdated nightly image

* test new fatimage build

* debug dnf remove cockpit

* --amend

* add cuda build back in

* cuda nightly build

* test cuda nightly builds

* test new fatimage build on SMS

* test image upploads across clouds

* test image uploads in separate workflow

* finish nightly build workflow

* fix image delete logic

* use azimuth-cloud trivy db mirror

* use GITHUB_TOKEN env

* test new fatimage build

* add final nightlybuilds workflow

* move trivy scan to separate workflow

* bump image and test new trivy scan

* fix artifact creation

* bump image and test trivy scan

* only run trivy scan on image bumps

* bump image to test trivy scan run condition

* bump cuda image

* bump image

* extend timeout for trivy scanning cuda image

* Run workflow on PR to main

* address PR comments

* fix source_image_name packer parse

* bump image

* additional PR comments

* bump image

Co-authored-by: Scott Davidson <[email protected]>

* Update README.md

* OSes supported as deploy hosts

* undo readme OSes supported

* add operations docs

* simplify main README.md to only cover default configuration

* move more-specific documentation into their own files

* provide site docs directory

* address docs review comments

* Fix a / in docs

Co-authored-by: Scott Davidson <[email protected]>

* address PR comments on docs

* address PR comments on docs

---------

Co-authored-by: bertiethorpe <[email protected]>
Co-authored-by: Scott Davidson <[email protected]>