The latest articles on DEV Community by Ezequiel (@ezequielpa4). https://dev.to/ezequielpa4 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3931623%2F33ed1389-fcc6-456e-8323-fded27d7432f.jpeg https://dev.to/ezequielpa4 en Ezequiel Thu, 14 May 2026 16:14:33 +0000 https://dev.to/ezequielpa4/ceph-on-ubuntu-single-node-setup-guide-osd-iscsi-bonding-342 https://dev.to/ezequielpa4/ceph-on-ubuntu-single-node-setup-guide-osd-iscsi-bonding-342 <p>I recently put together a small Ceph lab on Ubuntu using cephadm, mainly to understand how everything fits together in a real environment.</p> <p>Most guides out there are either too abstract or assume multi-node production setups, so I decided to document a <strong>single-node deployment</strong> end-to-end — focusing on what actually matters when you're getting started.</p> <h2> What this includes </h2> <ul> <li>Ceph installation (cephadm)</li> <li>OSD provisioning from previously used disks</li> <li>Network bonding (802.3ad / LACP)</li> <li>iSCSI gateway setup</li> <li>LVM snapshots for rollback scenarios</li> </ul> <p>Everything is based on Ubuntu Server 22.04 and Ceph Squid.</p> <h2> Repo </h2> <p>Full step-by-step docs here:</p> <p>👉 <a href="https://github.com/EzequielPA4/ceph-infrastructure-docs" rel="noopener noreferrer">https://github.com/EzequielPA4/ceph-infrastructure-docs</a></p> <h2> Why this might help </h2> <p>If you're:</p> <ul> <li>testing Ceph in a lab</li> <li>learning how OSDs really work</li> <li>trying to avoid common cephadm issues</li> <li>dealing with reused disks (LVM leftovers, etc.)</li> </ul> <p>this will probably save you time.</p> <h2> Notes from the lab </h2> <p>A few things that are worth calling out:</p> <ul> <li>Ceph is <strong>very strict with disks</strong> → anything with existing LVM or FS will be rejected</li> <li>Cleaning disks properly is key (wipefs + sgdisk + dd in some cases)</li> <li>Bonding (LACP) needs correct switch config or it just won’t behave</li> <li>Snapshots with LVM are useful, but you need to monitor usage (<code>lvs -o +data_percent</code>)</li> </ul> <h2> Limitations </h2> <p>This is a <strong>single-node setup</strong>, so:</p> <ul> <li>no HA</li> <li>no quorum discussion</li> <li>focused on learning / testing</li> </ul> <h2> If you're running something similar </h2> <p>I'm curious how others are setting this up in lab environments.</p> <ul> <li>Are you using cephadm or something else?</li> <li>Any gotchas with iSCSI gateways?</li> </ul> <p>Feel free to share your setup or improvements.</p> ceph linux homelab devops