Describe the issue ================== Steps to configure stretched ceph cluster are bit confusing in some cases, and could be optimized and simplified. Describe the task you were trying to accomplish =============================================== Deploy a stretched ceph cluster for MetroDR. Suggestions for improvement =========================== Steps to have ceph installed (as presented in the chapter) should be summarized in a succint overview in the beggining of the chapter. Section "3.4. Node pre-deployment requirements" explains few manual steps to perform, but some of these are also performed during ephadm-preflight.yml (eg. setup of red hat channels or installation of podman ...), so there is no need to perform them on all nodes manually, only to run ready to run playbook to do it again. There should be no overlap between the manual steps and the preflight playbook. Stress out that ceph1 is a deployment node, referencing ceph guide. It's not immediatelly clear whether steps in section 3.4 should be done to all nodes (then my suggestion above applies), or on a ceph1 only (but then one don't need to sprinkle "do this on ceph1 only" notes every now and then). Maybe we should cover setup of a deployment node in a separate section or step. I would avoid listing steps to configure hostnames via /etc/hosts file, since this makes the text longer, while we can assume readers can configure it via hosts files or DNS as it finds suitable in their environment. Instead, note what one needs to achieve and how to quickly check it, so that it applies to all environments. We should not assume that ec2-user is used on the machines, since most of the production deployments will be running on premise. We could use a different account name instead if we insist on having an example with non root access. Document URL ============ https://access.redhat.com/documentation/en-us/red_hat_openshift_data_foundation/4.10/html/configuring_openshift_data_foundation_for_metro-dr_with_advanced_cluster_management/requirements-for-deploying-rhcs-stretch-cluster_rhodf Chapter/Section Number and Title ================================ Chapter 3. Requirements for deploying Red Hat Ceph Storage stretch cluster with arbiter Product Version =============== ODF 4.10 Environment Details =================== MetroDR with ACM: deployment with stretched RHCS cluster. Any other versions of this document that also needs this update =============================================================== This text was introduced in 4.10, and should be fixed at least in 4.11. Additional information ====================== See also other doc BZs in MetroDR, which covers dedicated topics.
Hi Olive, added a comment in gdoc section: 4.5.2 to address comment 10 made by Martin in this BZ. With that modification the only pending item in this BZ would be: "summary of the steps: missing - I'm ok with not having this if you disagree for some reason (with the current shorter polished text, the need for a clear overview is not that strong)" This is something Olive you can take a look at and see if a short overview of the steps to deploy RHCS would fit or if with the doc TOC it's enough?.
Re-targeting the verification part
Checking preview build (commit bdf21e3): - [X] summary of the steps: ceph setup is now placed in a separate section "4.5. Deploying Red Hat Ceph Storage" with 3 subsections so that it's structure explains it well - [X] confusion about ceph1 only steps vs all nodes is resolved - [X] hosts file changes were reintroduced - [X] use --ssh-private-key and --ssh-public-key options with cephadm command >>> VERIFIED