A few weeks ago, I posted a blog about how SUSE pulled the plug on SUSE Enterprise Storage (SES). SES is a Linux-based computer data storage product developed by SUSE and built on Ceph technology. A quick recap on the matter is that in 2020, SUSE acquired Rancher Labs, a Kubernetes management software vendor, and short afterwards decided to stop the support of SES. Existing customers of SES will receive limited support until the end of 2022 and SES 6 support will end in January 2022. Also, from now on, no new clients will be accepted for SES and there will not be any further updates. However, I have already covered this in our previous blog on SES, so, for this blog I will go a little more in depth about the possible options and steps to migrate from SES to upstream Ceph supported by 42on.
Because SUSE Enterprise Storage is a much used product based on Ceph, we got a lot of questions from SES users about possible follow up steps. Mostly by SES users who want to keep working with the same storage technology, which means they want to migrate to the open source version in combination with training, consultancy, and support.
At this moment some migration path options are:
1. Upgrade from SES6 to SES7, from here on migrate to Ceph Octopus and then migrate the operating system from SUSE Linux Enterprise Server (SLES) to Ubuntu.
2. Upgrade from SES6 to SES7 to Ubuntu and then to Ceph Octopus. The disadvantage in comparison with the first migration path is that a rollback scenario is not possible.
3. The third option is to migrate from SES6 to Ubuntu, from here to Ceph Nautilus and finally to Ceph Octopus. With this method Ceph Ansible can be used. In this way you minimize the change to miss any patches in Ceph which makes it the preferable way for larger Ceph clusters.
4. A fourth option is to move from SES6 to Ceph Nautilus, then from here to Ceph Octopus and finally to Ubuntu. Although this way is not preferable.
These paths make the base operating system deviate significantly from the default deployment generally. This means that the management of these systems will not be included in the organizations standard operating procedures. However, another option exists. we see that in practice a lot of SUSE Linux Enterprise Server (SLES) users are satisfied with the Operating System and preferably do not want to change to Ubuntu, so in this case users can migrate in the following way:
5. From SES6 to Containerized Ceph Octopus and then to SLES 15.1 as operating system.
By adding containerization and using cephadm, users decouple Ceph from the operating system. cephadm deploys and manages the Ceph cluster. It does this by connecting the manager daemon to hosts via SSH. The manager daemon is able to add, remove, and update Ceph containers. cephadm does not rely on external configuration tools such as Ansible, Rook, and Salt.
Cephadm manages the full lifecycle of a Ceph cluster. This lifecycle starts with the bootstrapping process, when cephadm creates a tiny Ceph cluster on a single node. This cluster consists of one monitor and one manager. cephadm then uses the orchestration interface (“day 2” commands) to expand the cluster, adding all hosts and provisioning all Ceph daemons and services. Management of this lifecycle can be performed either via the Ceph command-line interface (CLI) or via the dashboard (GUI).
By using Cephadm, users have far more freedom and flexibility to think about the future migration paths. The only bottleneck of cephadm is the management of hosts when deployed in larger Ceph environments (we advise no more than 20 to 30 node deployments). Although this bottleneck is set to be solved in a future release.
So, summarized: SES users with more than 20 to 30 nodes, who want to migrate from SES to Ceph, might want to look at migration path 3 using Ceph Ansible. Although SES users with smaller SES clusters, working with SUSE and who want to continue using SUSE as base operating system (SLES), might want to look at migration path 5 using cephadm.
All described migration paths are supported by 42on and can include the migration from SUSE SES to upstream Ceph, Ceph training and the migration of the SUSE support services to 42on Ceph support services. As we already have experience with these migrations, we have created a structured migration path and plan for each option. With these paths we have created and our team’s expertise, we can work together with your teams to finish the migration within a couple of days.
Are you migrating from SES to Ceph as well? Or are you migrating to another solution? Will you be using similar migration paths? I am very curious to hear about your plans!