osd kerying用cli创建

ceph auth get-or-create osd.222
ceph auth caps osd.222 mgr 'allow profile osd' mon 'allow profile osd' osd 'allow *

手动创建配置文件导入

[osd.222]
key = AQBm7CxhMVXjJxAA7aoY+cZcG9aK7qLoU4X2Rw==
caps mon = "allow profile osd"
caps mgr = "allow profile osd"
caps osd = "allow *"

删除osd

ceph osd out 36
ceph osd stop 36
ceph osd rm 36
ceph auth rm osd.36
ceph osd crush remove osd.36

手动创建osd

ceph-volume prepare /dev/sdb
ceph-volume lvm activate osd.id fsid ceph-volume lvm activate 2 08956dbe-0b6e-4577-8b00-ada197a05ac9

设置pool允许删除

ceph config set mon mon_allow_pool_delete true

查看所有key

ceph config-key ls

倒入镜像

qemu-img convert -p -f qcow2  -O raw ./0000_demo_winxp.qcow2 rbd:vm/0000_demo_winxp

倒出镜像

qemu-img convert -p -f raw -O qcow2 rbd:vm/9ed73e48-f451-4dae-a10d-912de83dd028  ./xx_.qcow2

映射块设备到dev并挂载

rbd map 083b9e72-7653-4615-a09a-98487529b938 --pool v_disk
rbd showmapped   
mount /dev/rbd0 /mnt/data  
rbd unmap pool/images

podman 删除某个实例

cephadm ls #查看实例  
cephadm rm-daemon -n mds.cephfs-bigdata --fsid 86150a92-c060-11eb-bec5-bc97e1b56fd1

创建文件系统

ceph osd pool create cephfs_data 1024 1024 hdd_rule #创建pool 制定规则  
ceph osd pool create cephfs_metadata 1024 1024 hdd_rule  
ceph fs new cephfs cephfs_metadata cephfs_data  
ceph orch apply mds cephfs #创建mds

挂载文件系统

mount -t ceph p2:6789,p5:6789:/ /mnt/cephfs/ -o fs=fs name=admin

cursh操作

提取CRUSH Map  
ceph osd getcrushmap -o /tmp/crush

反编译crush图  
crushtool -d /tmp/crush -o /tmp/decompiled_crush

反编译crush图  
crushtool -c /tmp/decompiled_crush -o /tmp/crush_new

注入   
ceph osd setcrushmap -i /tmp/crush_new