#ceph
-
How to Add an OSD to Ceph Safely with cephadm
Add a Ceph OSD without wiping the wrong disk: verify health, inventory the device, dry-run an OSD spec, deploy it, and watch the rebalance.
-
Ceph vs Gluster: Which to Choose for Your Cluster
Compare Ceph and GlusterFS by storage model, workload fit, hardware requirements, cluster size, and support outlook to choose the right platform.
-
Ceph Erasure Coding vs Replication: How to Choose
How erasure coding and replication differ in space amplification, failure tolerance and write cost, with a k+m overhead table and a rule for picking one.
-
Ceph Hardware Requirements: CPU, RAM, Disks, Network
Ceph hardware requirements: 3 hosts for replicated pools; EC needs k+m. Compare 1-4 CPU threads, 4 GiB RAM per OSD, drives, and 10 Gb/s networking.
-
Ceph HEALTH_WARN: Diagnose and Clear Warnings
A triage order for Ceph HEALTH_WARN covering degraded placement groups, nearfull OSDs, missed deep scrubs, autoscaler alerts and monitor clock skew.
-
How Ceph Places Data: CRUSH, Pools, and Placement Groups
An orientation to Ceph's data placement model, covering OSDs, pools, placement groups, CRUSH failure domains, and replication versus erasure coding.