[QA] Ceph-deploy 安装 Ceph 环境报错分析

报错和分析

[ceph_deploy][ERROR ] RuntimeError: NoSectionError: No section: ‘ceph’

Operation:    ceph-deploy install ceph1 ceph2 ceph3ERROR:    [ceph1][INFO  ] Running command: rpm -Uvh --replacepkgs https://download.ceph.com/rpm-jewel/el7/noarch/ceph-release-1-0.el7.noarch.rpm    ......    [ceph_deploy][ERROR ] RuntimeError: NoSectionError: No section: 'ceph'Fix:    安装包冲突，直接移除已经安装的 rpm（把各个节点的这个包都移除）    yum remove ceph-release -y

[ceph_deploy][ERROR ] RuntimeError: Failed to execute command: ceph –version

Operation:    ceph-deploy install ceph1 ceph2 ceph3ERROR:    [ceph1][INFO  ] Running command: yum -y install ceph ceph-radosgw    ......    [ceph1][WARNIN] No data was received after 300 seconds, disconnecting...    [ceph1][INFO  ] Running command: ceph --version    ......    [ceph_deploy][ERROR ] RuntimeError: Failed to execute command: ceph --versionFix:    ceph1 安装速度过慢，已经超时了，直接手动安装    yum -y install ceph ceph-radosgw

[ceph1][ERROR ] “ceph auth get-or-create for keytype admin returned 1

Operation:    ceph-deploy mon create-initialERROR:    [ceph1][INFO  ] Running command: /usr/bin/ceph --connect-timeout=25 --cluster=ceph --admin-daemon=/var/run/ceph/ceph-mon.ceph1.asok mon_status    [ceph1][INFO  ] Running command: /usr/bin/ceph --connect-timeout=25 --cluster=ceph --name mon. --keyring=/var/lib/ceph/mon/ceph-ceph1/keyring auth get client.admin    [ceph1][ERROR ] "ceph auth get-or-create for keytype admin returned 1    [ceph1][DEBUG ] 2017-05-17 16:23:57.333371 7fafc0106700  0 -- :/1111132495 >> 192.168.122.18:6789/0 pipe(0x7fafbc061830 sd=6 :0 s=1 pgs=0 cs=0 l=1 c=0x7fafbc05d6e0).fault    ......    [ceph1][DEBUG ] Traceback (most recent call last):    [ceph1][DEBUG ]   File "/usr/bin/ceph", line 948, in <module>    [ceph1][DEBUG ]     retval = main()    [ceph1][DEBUG ]   File "/usr/bin/ceph", line 852, in main    [ceph1][DEBUG ]     prefix='get_command_descriptions')    [ceph1][DEBUG ]   File "/usr/lib/python2.7/site-packages/ceph_argparse.py", line 1300, in json_command    [ceph1][DEBUG ]     raise RuntimeError('"{0}": exception {1}'.format(argdict, e))    [ceph1][DEBUG ] RuntimeError: "None": exception "['{"prefix": "get_command_descriptions"}']": exception You cannot perform that operation on a Rados object in state configuring.    [ceph1][ERROR ] Failed to return 'admin' key from host ceph1    [ceph_deploy.gatherkeys][ERROR ] Failed to connect to host:ceph1    [ceph_deploy.gatherkeys][INFO  ] Destroy temp directory /tmp/tmpWajR2W    [ceph_deploy][ERROR ] RuntimeError: Failed to connect any monFix:    ceph.conf 中 public network 配置错误，和 mon_host 不在同一个网段。    配置正确的应该是：        mon_host = 192.168.122.18        public network = 192.168.122.18/24    之前的配置是：        mon_host = 192.168.122.18        public network = 172.16.34.253/24    这边是有点意思的地方，ceph-deploy 启动 monitor 的时候会优先使用 public network 配置的网络（ceph1 的 172.16.34.253）启动监听端口（6789），但是在和 monitor 连接的时候使用的却是 mon_host 的 IP(即 192.168.122.18)，会一致连接 192.168.122.18:6789，很显然联不通。这个就是错误的原因。虽然是一个很低级的错误，但是这个却让人感觉挺奇怪的。