meta data for this page
Differences
This shows you the differences between two versions of the page.
| Both sides previous revisionPrevious revisionNext revision | Previous revision | ||
| vm:proxmox:ceph:performance [2026/04/01 17:31] – niziak | vm:proxmox:ceph:performance [2026/06/07 21:13] (current) – niziak | ||
|---|---|---|---|
| Line 13: | Line 13: | ||
| Ceph is build for scale and works great in large clusters. In small cluster every node will be heavily loaded. | Ceph is build for scale and works great in large clusters. In small cluster every node will be heavily loaded. | ||
| + | * ceph ensure data safety - it waits for data to be written to medium on all replicas. Use enterpise SSDs with battery PLP (Power Loss Protection) to reduce latency. Some people reports 8x speed increase. | ||
| * adapt PG to number of OSDs to spread traffic evenly | * adapt PG to number of OSDs to spread traffic evenly | ||
| * use '' | * use '' | ||
| Line 19: | Line 20: | ||
| * MTU 9000 (jumbo frames) [[https:// | * MTU 9000 (jumbo frames) [[https:// | ||
| * net latency <200us ('' | * net latency <200us ('' | ||
| - | * [[https:// | + | * C-States: |
| * Ceph is incredibly sensitive to latency introduced by CPU c-state transitions. Set '' | * Ceph is incredibly sensitive to latency introduced by CPU c-state transitions. Set '' | ||
| * Disable IOMMU in kernel | * Disable IOMMU in kernel | ||