高可用OpenStack（Queen版）集群-15.Glance&Cinder集成Ceph

Posted 2020-12-26 netonline

参考文档：

1. Install-guide：https://docs.openstack.org/install-guide/
2. OpenStack High Availability Guide：https://docs.openstack.org/ha-guide/index.html
3. 理解Pacemaker：http://www.cnblogs.com/sammyliu/p/5025362.html
4. Ceph: http://docs.ceph.com/docs/master/start/intro/

十九．Glance集成Ceph

1. 配置glance-api.conf

# 在运行glance-api服务的节点修改glance-api.conf文件，含3个控制节点，以controller01节点为例
# 以下只列出涉及glance集成ceph的section
[[email protected] ~]# vim /etc/glance/glance-api.conf
# 打开copy-on-write功能
[DEFAULT]
show_image_direct_url = True

# 变更默认使用的本地文件存储为ceph rbd存储；
# 注意红色字体部分前后一致
[glance_store]
#stores = file,http
#default_store = file
#filesystem_store_datadir = /var/lib/glance/images/
stores = rbd
default_store = rbd
rbd_store_chunk_size = 8
rbd_store_pool = images
rbd_store_user = glance
rbd_store_ceph_conf = /etc/ceph/ceph.conf

# 变更配置文件，重启服务
[[email protected] ~]# systemctl restart openstack-glance-api.service
[[email protected] ~]# systemctl restart openstack-glance-registry.service

2. 上传镜像

# 镜像上传后，默认地址为ceph集群（ID）的images pool下
[[email protected] ~]# openstack image create "cirros-qcow2" 
 --file ~/cirros-0.3.5-x86_64-disk.img  --disk-format qcow2 --container-format bare  --public

# 检查
[[email protected] ~]# rbd ls images

3. 定义pool类型

# images启用后，ceph集群状态变为：HEALTH_WARN
[[email protected] ~]# ceph -s

# 使用”ceph health detail”，能给出解决办法；
# 未定义pool池类型，可定义为‘cephfs‘, ‘rbd‘, ‘rgw‘等
[[email protected] ~]# ceph health detail

# 同时解决volumes与vms两个pool的问题
[[email protected] ~]# ceph osd pool application enable images rbd
[[email protected]01 ~]# ceph osd pool application enable volumes rbd
[[email protected] ~]# ceph osd pool application enable vms rbd

# 查看
[[email protected] ~]# ceph health detail
[[email protected] ~]# ceph osd pool application get images
[[email protected] ~]# ceph osd pool application get volumes
[[email protected] ~]# ceph osd pool application get vms

二十．Cinder集成Ceph

1. 配置cinder.conf

# cinder利用插件式结构，支持同时使用多种后端存储；
# 在cinder-volume所在节点设置cinder.conf中设置相应的ceph rbd驱动即可；
# 含3个计算（存储）节点，以compute01节点为例；
# 以下只列出涉及cinder集成ceph的section
[[email protected] ~]# vim /etc/cinder/cinder.conf
# 后端使用ceph存储
[DEFAULT]
enabled_backends = ceph

# 新增[ceph] section；
# 注意红色字体部分前后一致
[ceph]
# ceph rbd驱动
volume_driver = cinder.volume.drivers.rbd.RBDDriver
rbd_pool = volumes
rbd_ceph_conf = /etc/ceph/ceph.conf
rbd_flatten_volume_from_snapshot = false
rbd_max_clone_depth = 5
rbd_store_chunk_size = 4
rados_connect_timeout = -1
# 如果配置多后端，则“glance_api_version”必须配置在[DEFAULT] section
glance_api_version = 2
rbd_user = cinder
rbd_secret_uuid = 10744136-583f-4a9c-ae30-9bfb3515526b
volume_backend_name = ceph

# 变更配置文件，重启服务
[[email protected] ~]# systemctl restart openstack-glance-api.service
[[email protected] ~]# systemctl restart openstack-glance-registry.service

2. 验证

# 查看cinder服务状态，cinder-volume集成ceph后，状态”up”；
# 或：cinder service-list
[[email protected] ~]# openstack volume service list

3. 生成volume

1）设置卷类型

# 在控制节点为cinder的ceph后端存储创建对应的type，在配置多存储后端时可区分类型；
# 可通过“cinder type-list”查看
[[email protected] ~]# cinder type-create ceph

# 为ceph type设置扩展规格，键值” volume_backend_name”，value值”ceph”
[[email protected] ~]# cinder type-key ceph set volume_backend_name=ceph
[[email protected] ~]# cinder extra-specs-list

2）生成volume

# 生成volume；
# 最后的数字”1”代表容量为1G
[[email protected] ~]# cinder create --volume-type ceph --name ceph-volume1 1

# 检查生成的volume；
# 或：cinder list
[[email protected] ~]# openstack volume list

# 检查ceph集群的volumes pool
[[email protected] ~]# rbd ls volumes