[转] 关于 Ceph PG

本系列文章会深入研究 Ceph 以及 Ceph 和 OpenStack 的集成：

（1）安装和部署

（2）Ceph RBD 接口和工具

（3）Ceph 物理和逻辑结构

（4）Ceph 的基础数据结构

（5）Ceph 与 OpenStack 集成的实现

（6）QEMU-KVM 和 Ceph RBD 的 缓存机制总结

（7）Ceph 的基本操作和常见故障排除方法

（8）关于Ceph PGs

https://docs.google.com/presentation/d/1_0eIWNYWvJON4GrJwktQ1Rumgqesc8QQbVkhutXgcNs/edit#slide=id.g7365a7bef_0_17

Placement Group, PG

How PGs are Used?

Should OSD#2 fail

Data Durability Issue

Object Distribution Issue

Choosing the number of PGs

Example

如何根據現狀調整 PG and PGP?

Monitoring OSDs

Ceph is NOT Healthy_OK

OSD Status Check

PG Sets

When A OSD in the Acting Set is down

Up Set

Check PG Status

Point

Peering

Peering: Establish Agreement of the PG status

Monitoring PG States

Check PG Stat

List Pool

PG IDs

The Output Format of the placement group

 PG

Creating PG

Create A Pool

 Peering

A Peering Process for a Pool with Replica 3

Active

Clean

DEGRADED

PG with {active + degraded}

Recovering

Backing Filling (1/2)

Backing Filling (2/2)

Remapped

 Stale

Identifying Troubled PGs (1/2)

Identify Trouble PGs (2/2)

Finding An Object Location