ATS-IT Consolidation of Server Provisioning
Zoom
CERN
Meeting 16/03/2026 11h00 zoom
ATS -IT server consolidation
Meeting took place on Zoom
present
IT) Jose / Kuba / Nikos
ATS) Thomas / Giuseppe / Enzo
Status of ATS-IT cloud POC:
IT deliver a complete RUNDECK WEB GUI dedicated API to ATS to provide delegation of HW JOBS
logs are stored in /var/log of the srv ssh -i with the provided keys.
>Create job for the
Servicing => alarm when
Then try to evacuate all VM nodes live, then put the HW in (maintenance) wait and then repair or upgrade them.
While in maintenance you can evacuate and repair boot any image run hided jobs and after validation, they are (hypervisor/BareMetal)
New feature would be useful for ironic for detailing and listing all actions and then applying then in one shot.
>ATS Repurpose use case. All nodes are stored in information into 3 DB.
The couple of important “Set management / alias / IP”
Jose Leon would try to define a way of proceeding.
FLAG to remove and decommission nodes, monitor the state and wait until . > Wait IT green light!
OS Image is targeting most common images.
-FW upgrades on BMC/IPMI, based on first check then only on request!
/!\ Project MATTERMOST Created by Kuba.
Next steps:
Check the network Stresstest/ burning test
1 node is able to be deployed and check can be used!
Copy of RHEL9 image where local image service aims2
>Enzo will provides a NFS place for OS images!
=> Meeting end 12:02
CHAT:
toulevey to Everyone 11:07 AM
https://gitlab.cern.ch/acc-adm/virt/packer-centos-kvm
Kuba to Everyone 11:17 AM
http://gitlab.cern.ch/cloud-infrastructure/ironic/-/blob/cern/2025.1/ironic/drivers/modules/cern/aims.py?ref_type=heads#L134-L135
https://linux.web.cern.ch/rhel/locmap/
Jose Castro Leon to Everyone 11:32 AM
https://gitlab.cern.ch/cloud-infrastructure/rundeck-bash/-/raw/master/Cloud-Operations/Delivery_provision/delivery_provision.sh
Jose Castro Leon to Everyone 11:46 AM
# OS_CLOUD=ironic_tn openstack baremetal node list --long -c Uuid -c 'Conductor Group'
+-----------------+--------------------------------------+
| Conductor Group | UUID |
+-----------------+--------------------------------------+
| tn_group_001 | b0dd79dc-713b-4256-8d47-557601f17492 |
| tn_group_002 | dbf1f004-1371-4529-b198-0a175e724bd9 |
| tn_group_002 | cbb18df8-2a4b-4b1f-bb5a-3198bd053e0d |
| tn_group_000 | 462784e2-c3a1-480a-b5d5-b622e22cd75d |
+-----------------+--------------------------------------+
- Conductor Group: tn_group_001
Name: dl7883651-mc19100021-1
- Conductor Group: tn_group_002
Name: dl7883651-mc19100021-3
- Conductor Group: tn_group_002
Name: dl7883651-mc19100021-4
- Conductor Group: tn_group_000
Name: dl7883651-mc19100021-2
Booting and ATS image specifics inside cloud region> need to copy to AIMS
RAID config image ?
Able to use aims2
Kerbeyros isn’t active in the image
Use rundeck must have a ATS job section ,
The Openstack HORIZON WEB API.
CCR nodes fluenbit.
CCR Server Console.
IT would have a very tight agenda after Q2 and would like!
NIKOS: PXE image run a script auto-registration flags per delivery point with script and review it for restAPI switchinfo+portinfo.
Gitlab is a dependency
Phase 2 dependency could be reviewed and adapted in case of needs
Next step fix RUNDECK