kolla-ansible安装cinder

LVM后端

环境拓扑

节点 IP 主机名
Controller/Network/Apollo 92.0.0.11 anode
Compute/Storage 92.0.0.12 bnode

multinode配置如下

[control]
92.0.0.11

[network]
92.0.0.11

[compute]
92.0.0.12

[monitoring]
92.0.0.11

[storage]
92.0.0.12

[apollo]
92.0.0.11

[extend]
92.0.0.10

Storage节点配置

新增磁盘(虚拟机直接添加磁盘、物理机使用losetup模拟),然后创建物理卷PV

# pvcreate /dev/sda
# pvdisplay
  --- Physical volume ---
  PV Name               /dev/vda2
  VG Name               centos
  PV Size               99.51 GiB / not usable 3.00 MiB
  Allocatable           yes 
  PE Size               4.00 MiB
  Total PE              25474
  Free PE               16
  Allocated PE          25458
  PV UUID               riLhOd-VTF3-soV7-2mqC-Epk9-zAM6-zQZHqg

  --- Physical volume ---
  PV Name               /dev/sda
  VG Name               cinder-volumes
  PV Size               10.00 GiB / not usable 4.00 MiB
  Allocatable           yes
  PE Size               4.00 MiB
  Total PE              2559
  Free PE               2559
  Allocated PE          0
  PV UUID               qs5MVX-c567-Kl7w-59uQ-1f7l-ufxv-mOBy8n

在新创建的物理卷上创建cinder-volume卷组

# vgcreate cinder-volumes /dev/sda
# vgdisplay 
  --- Volume group ---
  VG Name               centos
  System ID
  Format                lvm2
  Metadata Areas        1
  Metadata Sequence No  4
  VG Access             read/write
  VG Status             resizable
  MAX LV                0
  Cur LV                3
  Open LV               3
  Max PV                0
  Cur PV                1
  Act PV                1
  VG Size               99.51 GiB
  PE Size               4.00 MiB
  Total PE              25474
  Alloc PE / Size       25458 / 99.45 GiB
  Free  PE / Size       16 / 64.00 MiB
  VG UUID               neZcsa-uyAa-Qh1U-0RST-vMKw-fizT-9mhdlC

  --- Volume group ---
  VG Name               cinder-volumes
  System ID
  Format                lvm2
  Metadata Areas        1
  Metadata Sequence No  20
  VG Access             read/write
  VG Status             resizable
  MAX LV                0
  Cur LV                0
  Open LV               0
  Max PV                0
  Cur PV                1
  Act PV                1
  VG Size               10.00 GiB
  PE Size               4.00 MiB
  Total PE              2559
  Alloc PE / Size       0 / 0
  Free  PE / Size       2559 / 10.00 GiB
  VG UUID               nVNcsk-SU7J-vmfo-4blS-HiHS-mdeU-22ukmP

Kolla-ansible配置

设置配置文件

# vim /etc/kolla/globals.yml
enable_cinder: "yes"
enable_cinder_backend_lvm: "yes"

部署

# kolla-ansible -i multinode bootstrap-servers
# kolla-ansible -i multinode deploy

流程

### 创建卷流程
- cinder-volume服务使用"sudo cinder-rootwrap /etc/cinder/rootwrap.conf env LC_ALL=C lvcreate -n volume-03753745-933c-4f53-9e36-e90f13963aa5 cinder-volumes -L 1g"命令在cinder-volumes卷组上创建一个逻辑卷LV

### 卷挂载流程
- cinder-volume服务首先创建卷的iscsi target配置文件"/var/lib/cinder/volumes/volume-03753745-933c-4f53-9e36-e90f13963aa5",文件内如如下:
<target iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5>
    backing-store /dev/cinder-volumes/volume-03753745-933c-4f53-9e36-e90f13963aa5
    driver iscsi
    incominguser VbUHhzgSqYxDEdo8rtU5 hxsbZiw3HaSWc55h
    write-cache on
</target>

- 然后cinder-volume服务使用"sudo cinder-rootwrap /etc/cinder/rootwrap.conf tgt-admin --update iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5"命令创建iscsi target(服务端配置完成)

- 首先nova-libvirt服务使用iscsi客户端命令"iscsiadm -m node -T iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5 -p 92.0.0.12:3260 --interface default --op new"创建iscsi node(连接目标)

- 然后使用如下命令设置连接验证
  - "iscsiadm -m node -T iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5 -p 92.0.0.12:3260 --op update -n node.session.auth.authmethod -v CHAP"
  - "iscsiadm -m node -T iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5 -p 92.0.0.12:3260 --op update -n node.session.auth.username -v VbUHhzgSqYxDEdo8rtU5"
  - "iscsiadm -m node -T iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5 -p 92.0.0.12:3260 --op update -n node.session.auth.password -v ***"

- 接着使用"iscsiadm -m node -T iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5 -p 92.0.0.12:3260 --login"命令连接iscsi target(客户端配置完成)

- 最后生成"attach device xml",然后使用virsh命令添加磁盘,xml文件内容如下:
<disk type="block" device="disk">
  <driver name="qemu" type="raw" cache="none" io="native"/>
  <source dev="/dev/disk/by-path/ip-92.0.0.12:3260-iscsi-iqn.2010-10.org.openstack:volume-03753745-933c-4f53-9e36-e90f13963aa5-lun-1"/>
  <target bus="virtio" dev="vdb"/>
  <serial>03753745-933c-4f53-9e36-e90f13963aa5</serial>
</disk>

问题处理

  • 卷挂载失败
### nova-compute.log日志报错
2018-02-10 20:37:26.158 5 ERROR oslo_messaging.rpc.server   File "/var/lib/kolla/venv/lib/python2.7/site-packages/os_brick/initiator/connectors/iscsi.py", line 405, in connect_volume
2018-02-10 20:37:26.158 5 ERROR oslo_messaging.rpc.server     connection_properties)
2018-02-10 20:37:26.158 5 ERROR oslo_messaging.rpc.server   File "/var/lib/kolla/venv/lib/python2.7/site-packages/os_brick/initiator/connectors/iscsi.py", line 236, in _get_potential_volume_paths
2018-02-10 20:37:26.158 5 ERROR oslo_messaging.rpc.server     raise exception.FailedISCSITargetPortalLogin(message=msg)
2018-02-10 20:37:26.158 5 ERROR oslo_messaging.rpc.server FailedISCSITargetPortalLogin: Could not login to any iSCSI portal.

进到nova_compute容器,执行iscsiadm -m discovery -t sendtargets -p 92.0.0.12命令无法发现目标,且发现iscsid容器服务异常,所以先验证下iscsi target服务是否有异常,停掉iscsid和nova_compute容器,然后在宿主机器上安装iscsi客户端

# yum install iscsi-initiator-utils.x86_64 -y

发现并连接iscsi target

# iscsiadm -m discovery -t sendtargets -p 92.0.0.12
# iscsiadm -m node -T iqn.2010-10.org.openstack:volume-17876f1f-d28d-4ecb-aa86-ec0b7545e20b -p 92.0.0.12 -l

发现是可以正常连接的,说明服务端不存在问题,然后我们来验证下客户端,首先启动iscsid容器,然后在容器中执行下面操作

# iscsiadm -m discovery -t sendtargets -p 92.0.0.12
# iscsiadm -m node -T iqn.2010-10.org.openstack:volume-17876f1f-d28d-4ecb-aa86-ec0b7545e20b -p 92.0.0.12 -l

发现也是没有问题的,我们再把nova_compute容器启动起来,然后在nova_compute容器中执行下面操作

# iscsiadm -m discovery -t sendtargets -p 92.0.0.12

会发现命令卡住,无显示结果,这个时候我们还没发断定是iscsid容器的问题,还是nova-compute容器的问题,所以先停掉iscsid容器,然后在宿主机上启动iscsid服务(放在前端运行,方便看日志输出),然后继续在nova_compute容器中执行下面命令

# iscsiadm -m discovery -t sendtargets -p 92.0.0.12

会发现iscsid服务发生异常,同时nova_compute容器中执行的命令再次卡住,这时候说明nova_compute容器存在问题,最后发现nova-compute容器的iscsi客户端很久没有更新,而iscsid容器的客户端是最新的,怀疑是版本问题导致的异常,更新nova-compute容器的iscsi客户端版本后异常得到解决

NFS后端

环境拓扑

节点 IP 主机名
Controller/Network/Apollo 92.0.0.11 anode
Compute/Storage 92.0.0.12 bnode
NFS 92.0.0.13 cnode

multinode配置如下

[control]
92.0.0.11

[network]
92.0.0.11

[compute]
92.0.0.12

[monitoring]
92.0.0.11

[storage]
92.0.0.12

[apollo]
92.0.0.11

[extend]
92.0.0.10

NFS节点配置

安装服务

# yum install nfs-utils -y

设置配置文件

### 注意不能使用root用户创建的目录,否则后面cinder-volume创建卷时会出现Permission denied
$ mkdir -p /home/kolla/nfs
# tee /etc/exports << EOF
/home/kolla/nfs 92.0.0.0/24(rw,sync,insecure,no_root_squash)
EOF

启动服务

# systemctl enable nfs
# systemctl start nfs

验证NFS系统

# showmount -e 92.0.0.13
Export list for 92.0.0.13:
/kolla_nfs 92.0.0.0/24

Kolla-ansible配置

设置配置文件

# mkdir -p /etc/kolla/config
# tee /etc/kolla/config/nfs_shares.j2 << EOF
92.0.0.13:/home/kolla/nfs
EOF

# vim /etc/kolla/globals.yml
enable_cinder: "yes"
enable_cinder_backend_nfs: "yes"

部署

# kolla-ansible -i multinode bootstrap-servers
# kolla-ansible -i multinode deploy

流程

### 挂载流程
- cinder-volume服务会在/var/lib/cinder下创建一个目录用来挂载
- 读取/etc/cinder/nfs_shares中的nfs服务器信息,然后使用"sudo cinder-rootwrap /etc/cinder/rootwrap.conf mount -t nfs -o vers=4,minorversion=1 92.0.0.13:/home/kolla/nfs /var/lib/cinder/mnt/dd5f51c5ed990be3eb68a9e9b6099eb4"命令将其挂载到上面的目录

### 创建卷流程
- 使用"sudo cinder-rootwrap /etc/cinder/rootwrap.conf truncate -s 1G /var/lib/cinder/mnt/dd5f51c5ed990be3eb68a9e9b6099eb4/volume-cd53f360-1a42-480d-a059-424ebeac6509"命令创建对应大小的文件

### 卷挂载流程
- nova-libvirt根据创建的卷文件生成"attach device xml",然后使用virsh命令添加磁盘,xml文件内容如下:
<disk type="file" device="disk">
  <driver name="qemu" type="raw" cache="none" io="native"/>
  <source file="/var/lib/nova/mnt/dd5f51c5ed990be3eb68a9e9b6099eb4/volume-cd53f360-1a42-480d-a059-424ebeac6509"/>
  <target bus="virtio" dev="vdb"/>
  <serial>cd53f360-1a42-480d-a059-424ebeac6509</serial>
</disk>

问题处理

  • 卷创建失败
### cinder-volume.log日志报错
"/usr/bin/truncate: cannot open `/var/lib/cinder/mnt/dd5f51c5ed990be3eb68a9e9b6099eb4/volume-e3e2c8e1-66fa-4dda-a7a7-9c68514d596a' for writing: Permission denied\n"

查看下目录权限

# docker exec -it cinder_volume bash
$ ls -la /var/lib/cinder/mnt
total 0
drwxr-xr-x 3 cinder kolla  45 Mar  1 04:52 .
drwxrwsr-x 5 cinder kolla 109 Mar  1 05:06 ..
drwxr-xr-x 2   root  root  56 Mar  1 05:04 dd5f51c5ed990be3eb68a9e9b6099eb4

### 发现目录是root权限,所以`truncate`命令会执行失败,这边产生的原因是我之前的nfs服务器的配置文件内容是:/kolla_nfs 92.0.0.0/24(rw,sync,no_root_squash),并且这个目录是使用root用户创建的,使用普通用户创建,然后修改配置文件,重启nfs服务,并修改cinder-volume对应的nfs_shares配置文件,并重启容器即可解决问题

ISCSI后端

目前仅支持Hitachi storage volume driver

参考资料

Kolla-Ansible文档
Volume drivers

posted @ 2018-05-25 10:28  银魔术师  阅读(1649)  评论(0编辑  收藏  举报