软件安装过程中，考虑到现在是一个实验环境，且也考虑到规模不大，还有，网络压力不会大，出于简单考虑，将各个节点的拓扑结构改了一下，主要体现在网络节点和控制节点并在了一起。在一个服务器上安装！ 到目前位置，我的这个平台，只有keystone，glance，neutron，dashboard以及nova几个服务。省出的那个服务器，打算也作为计算节点。所以，最新规划topo如下了:
绿色节点表示目前已经安装了openstack的模块软件，灰色部分，表示下一步即将安装的部分。
下面简要说下安装的过程，重点记录一下我在Centos7下的安装中遇到的问题和解决办法。我若没有说在那个节点上安装，就表示是在控制节点node0上。
1. 安装NTP，官网安装用的是chrony，我折腾了半天，没有配置好，最终选择Linux系统自带的NTPD.相关的配置，参考我的博文NTP配置。
2. 安装yum repo以及openstack的CLI安装包程序以及selinux相关的openstack插件。
1 yum install centos-release-openstack-liberty
上一步安装完毕后，记得要执行下面一步，并且一定，最好，将操作系统重启，我在这一步没有做重启，直接继续，遇到了不少的问题，安装得我都要抓狂。。。
1 yum upgrade
客户端命令行插件，selinux的openstack插件
1 yum install python-openstackclient 2 yum install openstack-selinux
3. SQL database安装。
1 yum install mariadb mariadb-server MySQL-python
相关的配置/etc/my.cnf.d/mariadb_openstack.cnf：
1 [mysqld] 2 datadir=/var/lib/maria 3 socket=/var/lib/maria/maria.sock 4 # Disabling symbolic-links is recommended to prevent assorted security risks 5 symbolic-links=0 6 # Settings user and group are ignored when systemd is used. 7 # If you need to run mysqld under a different user or group, 8 # customize your systemd unit file for mariadb according to the 9 # instructions in http://fedoraproject.org/wiki/Systemd 10 11 bind-address = localhost 12 default-storage-engine = innodb 13 innodb_file_per_table 14 collation-server = utf8_general_ci 15 init-connect = 'SET NAMES utf8' 16 character-set-server = utf8 17 18 [mysqld_safe] 19 log-error=/var/log/mariadb/mariadb.log 20 pid-file=/var/run/mariadb/mariadb.pid
启动mariadb，由于mariadb和mysql是完全兼容的，可以说如何使用mysql的习惯，在mariadb上完全可以用上。启动后登录，执行mysql_secure_installation进行root密码的配置。
1 systemctl enable mariadb.service 2 systemctl start mariadb.service
4。 MongoDB的安装与配置 （version: 2.9.11）
1 yum install mongodb-server mongodb
配置/etc/mongod.conf：
1 ## 2 ### Basic Defaults 3 ## 4 5 # Comma separated list of ip addresses to listen on (all local ips by default) 6 bind_ip = 127.0.0.1 7 8 # Specify port number (27017 by default) 9 #port = 27017 10 11 # Fork server process (false by default) 12 #fork = true 13 14 # Full path to pidfile (if not set, no pidfile is created) 15 pidfilepath = /var/run/mongodb/mongod.pid 16 17 # Log file to send write to instead of stdout - has to be a file, not directory 18 logpath = /var/log/mongodb/mongod.log 19 20 # Alternative directory for UNIX domain sockets (defaults to /tmp) 21 unixSocketPrefix = /var/run/mongodb 22 23 # Directory for datafiles (defaults to /data/db/) 24 dbpath = /var/lib/mongodb 25 26 # Enable/Disable journaling (journaling is on by default for 64 bit) 27 #journal = true 28 nojournal = true 29 30 ## 31 ### General options 32 ## 33 34 # Be more verbose (include multiple times for more verbosity e.g. -vvvvv) (v by default) 35 #verbose = v 36 37 # Max number of simultaneous connections (1000000 by default) 38 #maxConns = 1000000 39 40 # Log to system's syslog facility instead of file or stdout (false by default) 41 #syslog = true 42 43 # Syslog facility used for monogdb syslog message (user by defautl) 44 #syslogFacility = user 45 46 # Append to logpath instead of over-writing (false by default) 47 #logappend = true 48 49 # Desired format for timestamps in log messages (One of ctime, iso8601-utc or iso8601-local) (iso8601-local by default) 50 #timeStampFormat = arg 51 52 # Private key for cluster authentication 53 #keyFile = arg 54 55 # Set a configurable parameter 56 #setParameter = arg 57 58 # Enable http interface (false by default) 59 #httpinterface = true 60 61 # Authentication mode used for cluster authentication. Alternatives are (keyFile|sendKeyFile|sendX509|x509) (keyFile by default) 62 #clusterAuthMode = arg 63 64 # Disable listening on unix sockets (false by default) 65 #nounixsocket = true 66 67 # Run with/without security (without by default) 68 #auth = true 69 #noauth = true 70 71 # Enable IPv6 support (disabled by default) 72 #ipv6 = true 73 74 # Allow JSONP access via http (has security implications) (false by default) 75 #jsonp = true 76 77 # Turn on simple rest api (false by default) 78 #rest = true 79 80 # Value of slow for profile and console log (100 by default) 81 #slowms = 100 82 83 # 0=off 1=slow, 2=all (0 by default) 84 #profile = 0 85 86 # Periodically show cpu and iowait utilization (false by default) 87 #cpu = true 88 89 # Print some diagnostic system information (false by default) 90 #sysinfo = true 91 92 # Each database will be stored in a separate directory (false by default) 93 #directoryperdb = true 94 95 # Don't retry any index builds that were interrupted by shutdown (false by default) 96 #noIndexBuildRetry = true 97 98 # Disable data file preallocation - will often hurt performance (false by default) 99 #noprealloc = true 100 101 # .ns file size (in MB) for new databases (16 MB by default) 102 #nssize = 16 103 104 # Limits each database to a certain number of files (8 default) 105 #quota 106 107 # Number of files allowed per db, implies --quota (8 by default) 108 #quotaFiles = 8 109 110 # Use a smaller default file size (false by default) 111 #smallfiles = true 112 113 # Seconds between disk syncs (0=never, but not recommended) (60 by default) 114 #syncdelay = 60 115 116 # Upgrade db if needed (false by default) 117 #upgrade = true 118 119 # Run repair on all dbs (false by default) 120 #repair = true 121 122 # Root directory for repair files (defaults to dbpath) 123 #repairpath = arg 124 125 # Disable scripting engine (false by default) 126 #noscripting = true 127 128 # Do not allow table scans (false by default) 129 #notablescan = true 130 131 # Journal diagnostic options (0 by default) 132 #journalOptions = 0 133 134 # How often to group/batch commit (ms) (100 or 30 by default) 135 #journalCommitInterval = 100 136 137 138 139 ## 140 ### Replication options 141 ## 142 143 # Size to use (in MB) for replication op log (default 5% of disk space - i.e. large is good) 144 #oplogSize = arg 145 146 147 148 ## 149 ### Master/slave options (old; use replica sets instead) 150 ## 151 152 # Master mode 153 #master = true 154 155 # Slave mode 156 #slave = true 157 158 # When slave: specify master as <server:port> 159 #source = arg 160 161 # When slave: specify a single database to replicate 162 #only = arg 163 164 # Specify delay (in seconds) to be used when applying master ops to slave 165 #slavedelay = arg 166 167 # Automatically resync if slave data is stale 168 #autoresync = true 169 170 171 172 ## 173 ### Replica set options 174 ## 175 176 # Arg is <setname>[/<optionalseedhostlist>] 177 #replSet = arg 178 179 # Specify index prefetching behavior (if secondary) [none|_id_only|all] (all by default) 180 #replIndexPrefetch = all 181 182 183 184 ## 185 ### Sharding options 186 ## 187 188 # Declare this is a config db of a cluster (default port 27019; default dir /data/configdb) (false by default) 189 #configsvr = true 190 191 # Declare this is a shard db of a cluster (default port 27018) (false by default) 192 #shardsvr = true 193 194 195 196 ## 197 ### SSL options 198 ## 199 200 # Use ssl on configured ports 201 #sslOnNormalPorts = true 202 203 # Set the SSL operation mode (disabled|allowSSL|preferSSL|requireSSL) 204 # sslMode = arg 205 206 # PEM file for ssl 207 #sslPEMKeyFile = arg 208 209 # PEM file password 210 #sslPEMKeyPassword = arg 211 212 # Key file for internal SSL authentication 213 #sslClusterFile = arg 214 215 # Internal authentication key file password 216 #sslClusterPassword = arg 217 218 # Certificate Authority file for SSL 219 #sslCAFile = arg 220 221 # Certificate Revocation List file for SSL 222 #sslCRLFile = arg 223 224 # Allow client to connect without presenting a certificate 225 #sslWeakCertificateValidation = true 226 227 # Allow server certificates to provide non-matching hostnames 228 #sslAllowInvalidHostnames = true 229 230 # Allow connections to servers with invalid certificates 231 #sslAllowInvalidCertificates = true 232 233 # Activate FIPS 140-2 mode at startup 234 #sslFIPSMode = true
启动
1 systemctl enable mongod.service 2 systemctl start mongod.service
5. RabbitMQ安装
1 yum install rabbitmq-server
1 systemctl enable rabbitmq-server.service 2 systemctl start rabbitmq-server.service
创建用户openstack并修改其在vhost下的配置，读写权限
1 rabbitmqctl add_user openstack RABBIT_PASS
1 rabbitmqctl set_permissions openstack ".*" ".*" ".*"
我为了操作方便简单，易于记忆，将所有的和密码相关的信息，设置了同一个密码，都是openstack。
以上是一些基本的环境的准备，接下来，就是要安装具体的openstack的组成部件了。第一个是keystone的安装。
k1。 创建keystone数据库，设置访问权限（密码也是openstack，偷懒）
1 mysql -u root -p 2 CREATE DATABASE keystone; 3 GRANT ALL PRIVILEGES ON keystone.* TO 'keystone'@'localhost' IDENTIFIED BY 'KEYSTONE_DBPASS'; 4 GRANT ALL PRIVILEGES ON keystone.* TO 'keystone'@'%' IDENTIFIED BY 'KEYSTONE_DBPASS';
k2. 安装keystone，httpd，缓存以及wsgi并启动
1 yum install openstack-keystone httpd mod_wsgi memcached python-memcached 2 3 systemctl enable memcached.service 4 systemctl start memcached.service
配置/etc/keystone/keystone.conf:
1 [DEFAULT] 2 admin_token = 613ae457d94da5033c92 3 verbose = true 4 5 [database] 6 connection = mysql://keystone:openstack@192.168.1.100/keystone 7 8 [memcache] 9 servers = localhost:11211 10 11 [revoke] 12 driver = sql 13 14 [token] 15 provider = uuid 16 driver = memcache
接下来就是数据库keystone的同步操作，对于第一次操作这个命令，可能会遇到问题，就是No handlers could be found for logger "oslo_config.cfg"
1 su -s /bin/sh -c "keystone-manage db_sync" keystone
在我的部署过程中，这个无关大局。就让它在那吧。。
配置httpd /etc/httpd/conf/httpd.conf ，其他的都默认不改。就修改一下下面这行。
1 ServerName 192.168.1.100
配置wsgi /etc/httpd/conf.d/wsgi-keystone.conf：
1 Listen 5000 2 Listen 35357 3 4 <VirtualHost *:5000> 5 WSGIDaemonProcess keystone-public processes=5 threads=1 user=keystone group=keystone display-name=%{GROUP} 6 WSGIProcessGroup keystone-public 7 WSGIScriptAlias / /usr/bin/keystone-wsgi-public 8 WSGIApplicationGroup %{GLOBAL} 9 WSGIPassAuthorization On 10 <IfVersion >= 2.4> 11 ErrorLogFormat "%{cu}t %M" 12 </IfVersion> 13 ErrorLog /var/log/httpd/keystone-error.log 14 CustomLog /var/log/httpd/keystone-access.log combined 15 16 <Directory /usr/bin> 17 <IfVersion >= 2.4> 18 Require all granted 19 </IfVersion> 20 <IfVersion < 2.4> 21 Order allow,deny 22 Allow from all 23 </IfVersion> 24 </Directory> 25 </VirtualHost> 26 27 <VirtualHost *:35357> 28 WSGIDaemonProcess keystone-admin processes=5 threads=1 user=keystone group=keystone display-name=%{GROUP} 29 WSGIProcessGroup keystone-admin 30 WSGIScriptAlias / /usr/bin/keystone-wsgi-admin 31 WSGIApplicationGroup %{GLOBAL} 32 WSGIPassAuthorization On 33 <IfVersion >= 2.4> 34 ErrorLogFormat "%{cu}t %M" 35 </IfVersion> 36 ErrorLog /var/log/httpd/keystone-error.log 37 CustomLog /var/log/httpd/keystone-access.log combined 38 39 <Directory /usr/bin> 40 <IfVersion >= 2.4> 41 Require all granted 42 </IfVersion> 43 <IfVersion < 2.4> 44 Order allow,deny 45 Allow from all 46 </IfVersion> 47 </Directory> 48 </VirtualHost>
最后启动httpd，这个是支持后端rest api调用的服务器。
1 systemctl enable httpd.service 2 systemctl start httpd.service
k3. 创建keystone的service.首先配置环境变量，这里，我将这个简单版本的安装过程中涉及到的所有环境变量，都列在这里：
1 export MARIA_DB_ROOT_PW=openstack 2 export RABBIT_MQ_PASS=openstack 3 export KEYSTONE_DBPASS=openstack 4 export keystone_user_pw=openstack 5 6 export ADMIN_TOKEN=613ae457d94da5033c92 7 export OS_TOKEN=$ADMIN_TOKEN 8 export OS_URL=http://192.168.1.100:35357/v3 9 export OS_IDENTITY_API_VERSION=3 10 11 export GLANCE_DBPASS=openstack 12 export glance_user_pw=openstack 13 14 export NOVA_DBPASS=openstack 15 export nova_user_pw=openstack 16 17 export NEUTRON_DBPASS=openstack 18 export neutron_user_pw=openstack 19 20 export METADATA_SECRET=openstack
接下来，创建服务，以及对应的endpoint。
1 openstack service create --name keystone --description "OpenStack Identity" identity 2 openstack endpoint create --region RegionOne identity public http://node0:5000/v2.0 3 openstack endpoint create --region RegionOne identity internal http://node0:5000/v2.0 4 openstack endpoint create --region RegionOne identity admin http://node0:5000/v2.0
上面的第一步，创建keystone这个服务时，就遇到错误：
1 [root@node0 opt]# openstack service create --name keystone --description "OpenStack Identity" identity 2 Internal Server Error (HTTP 500)
其实，这个通过查看keystone的日志，注意，在httpd目录下哟，会发现访问数据库出错了。这个错误，是因为keystone-manage db_sync keystone没有做成功。我通过多次重新安装keystone组件，最终成功了。汗。。。
k4. 创建project admin以及用户以及角色
1 openstack project create --domain default --description "Admin Project" admin 2 openstack user create --domain default --password-prompt admin 3 openstack role create admin 4 openstack role add --project admin --user admin admin
k5. 创建project service
1 openstack project create --domain default --description "Service Project" service
k6. 创建project demo以及用户和角色
1 openstack project create --domain default --description "Demo Project" demo 2 openstack user create --domain default --password-prompt demo 3 openstack role create user 4 openstack role add --project demo --user demo user
k7. 验证keystone的安装正确性。
首先要修改/usr/share/keystone/keystone-dist-paste.ini，关闭临时token的auth机制。将下面红色的部分去掉。其实，在我的配置中，没有做这个操作，最终也可以达到创建instance，且可以操作新建的instance。【我将集群机器的防火墙都关闭了。。。】
1 [pipeline:public_api] 2 # The last item in this pipeline must be public_service or an equivalent 3 # application. It cannot be a filter. 4 pipeline = sizelimit url_normalize request_id build_auth_context token_auth admin_token_auth json_body ec2_extension user_crud_extension public_service 5 6 [pipeline:admin_api] 7 # The last item in this pipeline must be admin_service or an equivalent 8 # application. It cannot be a filter. 9 pipeline = sizelimit url_normalize request_id build_auth_context token_auth admin_token_auth json_body ec2_extension s3_extension crud_extension admin_service 10 11 [pipeline:api_v3] 12 # The last item in this pipeline must be service_v3 or an equivalent 13 # application. It cannot be a filter. 14 pipeline = sizelimit url_normalize request_id build_auth_context token_auth admin_token_auth json_body ec2_extension_v3 s3_extension simple_cert_extension revoke_extension federation_extension oauth1_extension endpoint_filter_extension service_v3
验证一下吧：
1 openstack --os-auth-url http://node0:35357/v3 --os-project-domain-id default --os-user-domain-id default --os-project-name admin --os-username admin --os-auth-type password token issue
是正确的，有数据显示，没有报错！通过demo这个用户操作，也是没有问题的。
k8. 为了操作时方便，将需要的参数source为环境变量吧。admin-openrc.sh/demo-openrc.sh,我这里，两个的文件内容一样。
1 #!/bin/bash 2 3 export OS_PROJECT_DOMAIN_ID=default 4 export OS_USER_DOMAIN_ID=default 5 export OS_PROJECT_NAME=admin 6 export OS_TENANT_NAME=admin 7 export OS_USERNAME=admin 8 export OS_PASSWORD=openstack 9 export OS_AUTH_TYPE=password 10 export OS_AUTH_URL=http://node0:35357/v3 11 export OS_IDENTITY_API_VERSION=3 12 13 #for glance 14 export OS_IMAGE_API_VERSION=2
这里，要注意了，将上面的第9行去掉，执行下面的命令，你会遇到一个奇怪的错误。
1 openstack token issue
错误如下：
1 [root@node0 opt]# openstack token issue 2 'NoneType' object has no attribute 'service_catalog'
呵呵，这个问题，我也google才找到原因的，就是官网的guide中少了那个OS_AUTH_TYPE。
到此，所有的keystone的基本配置都完成了。
开始安装image模块glance吧。很简单，这步！
g1. 创建数据库glance并建用户glance（密码还是openstack）
1 mysql -u root -p 2 CREATE DATABASE glance; 3 GRANT ALL PRIVILEGES ON glance.* TO 'glance'@'localhost' IDENTIFIED BY 'GLANCE_DBPASS'; 4 GRANT ALL PRIVILEGES ON glance.* TO 'glance'@'%' IDENTIFIED BY 'GLANCE_DBPASS';
接下来创建用户glance
1 openstack user create --domain default --password-prompt glance 2 openstack role add --project service --user glance admin
g2. 创建glance服务和endpoint
1 openstack service create --name glance --description "OpenStack Image service" image 2 openstack endpoint create --region RegionOne image public http://node0:9292 3 openstack endpoint create --region RegionOne image internal http://node0:9292 4 openstack endpoint create --region RegionOne image admin http://node0:9292
g3. 安装组件。
1 yum install openstack-glance python-glance python-glanceclient
g4. 配置/etc/glance/glance-api.conf以及/etc/glance/glance-registry.conf,两个的配置一样，其他的都采用默认值。
1 [DEFAULT] 2 notification_driver = noop 3 verbose = True 4 5 [database] 6 connection = mysql://glance:openstack@node0/glance 7 8 [keystone_authtoken] 9 auth_uri = http://node0:5000 10 auth_url = http://node0:35357 11 auth_plugin = password 12 project_domain_id = default 13 user_domain_id = default 14 project_name = service 15 username = glance 16 password = openstack 17 18 [paste_deploy] 19 flavor = keystone 20 21 [glance_store] 22 default_store = file 23 filesystem_store_datadir = /var/lib/glance/images/
g5. 数据库同步以及启动服务
1 su -s /bin/sh -c "glance-manage db_sync" glance 2 3 systemctl enable openstack-glance-api.service openstack-glance-registry.service 4 systemctl start openstack-glance-api.service openstack-glance-registry.service
g6. 验证。给glance上传一个cirros的镜像
1 wget http://download.cirros-cloud.net/0.3.4/cirros-0.3.4-x86_64-disk.img 2 3 glance image-create --name "cirros" --file cirros-0.3.4-x86_64-disk.img --disk-format qcow2 --container-format bare --visibility public --progress
同样，没有问题，很正常，glance image-list命令可以检测。
好了，今天就到这里吧，不早了，要回去收拾一下，要过年了。。。后面的部分，年后，我再将其补上。