Posts tagged with "ceph"

Nine of my fifteen Ceph OSDs use WAL-only acceleration while six use DB. I set out to measure the performance gap and discovered the real story isn't WAL vs DB — it's the USB 3.0 hardware ceiling that dominates everything. The matched-hardware comparison shows DB is 5-15% faster on reads, not the 32% that naive cross-node testing suggested.

Running Ceph on USB drives sounds crazy until you put the WAL and DB on an SSD. Here's how separating metadata onto a Crucial MX500 transformed my 15-OSD homelab cluster from sluggish to surprisingly capable — at a fraction of all-SSD costs.

A routine ZFS scrub alert on harlan turned into a multi-hour debugging session when a hostid mismatch fix collided with a pre-existing Ceph OSD failure from a dead USB drive. Here's how overlapping storage problems can mask each other and how to untangle them.

A power outage knocked my Ceph cluster from 15 healthy OSDs down to 4. The recovery took days of debugging — heartbeat cascades, a ceph.conf misconfiguration, and a dead SAN switch hiding behind NO-CARRIER flags on every node.

New Seagate USB drives arrived for the Ceph cluster and predictably won't report SMART data. After months of production use, GRUB boot parameters are the only method that reliably survives kernel updates and cluster reboots.

Real-world cost-per-GB analysis of distributed Ceph storage versus single-drive solutions in a homelab — the same infrastructure investment framework applies at any scale.

Performance tuning Ceph on USB storage and constrained hardware — mClock configuration, IOPS optimization, and the realities of USB 3.0 as a storage tier.

Ceph's nearfull warnings are capacity planning signals, not emergencies — if you understand the thresholds and respond before the cluster goes read-only.

Hard-won lessons from running Proxmox since version 7.4 through 8.2.2 — upgrade gotchas, Ceph integration tips, and the patterns that apply at any scale.

The Ceph Dashboard is essential for monitoring cluster health without SSH, but setting it up on Proxmox isn't straightforward. Here's how to get it working with SSL and proper authentication.

A publishing schedule for the remaining 16 articles planned through end of 2025, covering Proxmox, Ceph, hardware monitoring, CLI tools, and Jekyll optimization.

How to completely remove a broken Ceph cluster from Proxmox — every service, package, config file, and orphaned mount point — so you can start fresh.

Ceph Monitor refuses to start after a Proxmox 8.2.4 upgrade because root partitions hit 95% — fixed by clearing apt cache, removing atop logs, and purging old PVE and Debian kernels.

Tuning Ceph scrub and deep-scrub intervals to reduce wear on spinning rust drives in a homelab cluster — spreading daily scrubs over 7 days and weekly deep scrubs over 28 days.

Building a three-node Proxmox 8 + Ceph test cluster on Dell Wyse 3040 thin clients to safely evaluate SDN and Ceph configuration changes without risking the semi-production cluster.

Storage has hit a penny per GB. A 20TB renewed drive for $200 and a 20TB USB-C enclosure for $220 make the math work for a Ceph offsite backup strategy.

Building a Proxmox 8 + Ceph HA cluster on decade-old Dell OptiPlex 990 hardware with Seagate USB drives — the background, the hardware inventory, and why Ceph beat a Synology NAS.

Fixing a severely imbalanced Ceph cluster where OSDs added in batches left the first three drives at 75-85% while newer ones sat nearly empty — by tuning backfill and recovery parallelism.