Francesco,

Cephadm implements "rolling upgrade". It will stop the services one by one so that the cluster always remains healthy, with no impact on the services (except for rgw or MDS that may be single instance but it is just the time of a restart). And it will pause if at any point Moving forward may put the cluster at risk.

Michel
Sent from my mobile

Le 20 mai 2025 14:04:07 Francesco Di Nucci <francesco.dinucci@na.infn.it> a écrit :

Thanks a lot, this discussion is being very helpful!

-- 
Francesco Di Nucci
System Administrator 
Compute & Networking Service, INFN Naples

Email: francesco.dinucci@na.infn.it
On 20/05/25 10:57, Michel Jouvin wrote:
Hi Francesco,

We used to manage our Ceph cluster with a Puppet -like tool and decided to switch to cephadm to do the cluster manager in Octopus, using our system management tool only to provision the OS and Podman and configure the SSH key required by cephadm. It has been a huge improving terms of management efficiency, in particular the time required by any Ceph upgrade operation, including OS upgrades/réinstallation where cephadm redeploys everything required automatically.

We will never move back to previous way of managing our cluster!

Michel
Sent from my mobile

Le 20 mai 2025 09:33:07 Francesco Di Nucci <francesco.dinucci@na.infn.it> a écrit :

Thank you,

this might be a solution too (using other tools to setup the OS and then 
switch to cephadm)

It's not only about familiarity, I was also thinking about feasibility 
in the long run, with upgrades, new nodes etc, so nice to know that CEPH 
and OpenStack managements can be decoupled

Best regards

-- 
Francesco Di Nucci
System Administrator
Compute & Networking Service, INFN Naples


On 20/05/25 09:12, Eugen Block wrote:
Hi,

I would say this is opinion based and depends on your experience and 
infrastructure.
Even if you decided to use cephadm as a Ceph deployment tool, you 
still need to have some installation and configuration management in 
place, at least if you have more than a few hosts. Because with the 
Ceph orchestrator (cephadm) you can only add hosts that have been 
configured to your needs, like ssh keys, podman/docker, chrony etc. If 
you use puppet for that, it might be the right choice for you.
We are using a combination of cobbler and Salt (Uyuni project) to 
perform automatic OS installation via PXE boot and configuration via 
Salt. Once the systems are ready to join the Ceph cluster, we just add 
them via orchestrator (ceph orch host add ...) and then the rest is 
managed by cephadm. So in our case, Ceph is decoupled from OpenStack 
management, although the OpenStack hosts are also installed and 
configured via Salt.

I'd say choose the method you're most familiar with.

Regards,
Eugen

Zitat von Francesco Di Nucci <francesco.dinucci@na.infn.it>:

Thank you,

I'd read it, but as there also are other methods such as ceph-ansible 
and puppet-ceph I am trying to get a feedback from other operators 
about their experiences, as in this case I'm particularly interested 
in integration of CEPH with OpenStack

Best regards

-- 
Francesco Di Nucci
System Administrator
Compute & Networking Service, INFN Naples


On 19/05/25 16:29, Maksim Malchuk wrote:
Hi Francesco,

The CEPH community recommends using CEPHADM as the primary tool for 
deploying CEPH: 


On Mon, May 19, 2025 at 4:19 PM Francesco Di Nucci 

   Hi all,

   we're planning (finally) to setup a CEPH cluster to be used as
   OpenStack
   backend.

   The cloud is actually set up with Foreman+Puppet, to setup CEPH what
   would you advice?

   cephadm as it's the preferred method in the CEPH docs or Puppet
   with the
   puppet-ceph module, as it's part of openstack?

   Thanks in advance

   --     Francesco Di Nucci
   System Administrator
   Compute & Networking Service, INFN Naples





-- 
Regards,
Maksim Malchuk