onestack

#!/usr/bin/env bash
# **setup_base.sh** is a tool to deploy base and real OpenStack cloud computing service.
# This script installs and configures various combinations of *Glance*,
# *Horizon*, *Keystone*, *Nova*, *Mysql* and others.
# [email protected] (Kayven)
# Learn more and get the most recent version at http://code.google.com/p/onestack/
set -o xtrace
## 请使用root执行本脚本！
## Ubuntu 12.04 ("Precise") 部署 OpenStack Essex
## 参考：
## http://docs.openstack.org/essex/openstack-compute/starter/content/
## 本脚本建立基本的OpenStack系统，没有上传镜像，没有创建实例，使用 setup_test.sh可以完成这些。
## 注意：如果没有设置apt源、网络等，请先设置，或者使用oneStack.sh安装（该脚本里面有详细说明）。
## 一：准备系统
## 1、设置参数
##########################################################################
## 如果原来安装过OpenStack，请先执行 ./delStack.sh
##########################################################################
## 配置参数
MYSQL_PASSWD=${MYSQL_PASSWD:-"cloud1234"}
NOVA_DB_USERNAME=${NOVA_DB_USERNAME:-"novadbadmin"}
NOVA_DB_PASSWD=${NOVA_DB_PASSWD:-"cloud1234"}
GLANCE_DB_USERNAME=${GLANCE_DB_USERNAME:-"glancedbadmin"}
GLANCE_DB_PASSWD=${GLANCE_DB_PASSWD:-"cloud1234"}
## 自行检查下面network/interfaces的两个网卡设置
OUT_IP="192.168.139.50"
OUT_IP_PRE="192.168.139"
IN_IP="10.0.0.1"
IN_IP_PRE="10.0.0"
FLAT_IP="10.0.0.40"
FLOAT_IP="192.168.139.225"
## 选择虚拟技术，裸机使用kvm，虚拟机里面使用qemu
VIRT_TYPE="qemu"
## token, 登录dashboard密码
ADMIN_TOKEN="admin"
##########################################################################
# Determine what system we are running on.  This provides ``os_VENDOR``...
# Determine OS Vendor, Release and Update
#if [[ -x "`which lsb_release 2>/dev/null`" ]]; then
    os_VENDOR=$(lsb_release -i -s)
    os_RELEASE=$(lsb_release -r -s)
    os_UPDATE=""
    os_CODENAME=$(lsb_release -c -s)
#fi
if [ "Ubuntu" = "$os_VENDOR" ]; then
    DISTRO=$os_CODENAME
else
    echo "The os didn't seems to be Ubuntu."
    exit 1
fi
echo $DISTRO
if [ "precise" != ${DISTRO} -a "oneiric" != ${DISTRO} ]; then
    echo "WARNING: this script has been tested on oneiric or precise"
    exit 1
fi
############################################################################
############################################################################
## 4：设置网络
SOURCE_FILE=${SOURCE_FILE:-"/etc/apt/sources.list"}
# network configure
NETWORK_CONF=${NETWORK_CONF:-"/etc/network/interfaces"}
if ! grep -q eth1 $NETWORK_CONF; then
        cat <<INTERFACES >$NETWORK_CONF
auto lo
iface lo inet loopback
# The primary network interface
auto eth0
iface eth0 inet static
pre-up ifconfig eth0 hw ether b8:ac:6f:9a:ee:e4
        address 192.168.139.50
        netmask 255.255.255.0
        network 192.168.139.0
        broadcast 192.168.139.255
        gateway 192.168.139.253
        dns-nameservers 210.72.128.8
auto eth1
iface eth1 inet static
pre-up ifconfig eth1 hw ether b8:ac:6f:9a:ee:e4
        address 10.0.0.1
        netmask 255.255.255.0
        network 10.0.0.0
        broadcast 10.0.0.255
INTERFACES
        /etc/init.d/networking restart
fi
## 以上系统配置，可以自行配置，注释掉这些步骤。
############################################################################
############################################################################
apt-get update
## 5：安装bridge
apt-get install -y bridge-utils
## 6：设置NTP
apt-get install -y ntp
if ! grep -q fudge "/etc/ntp.conf"; then
        cat <<NTPconf > /etc/ntp.conf
server ntp.ubuntu.com iburst
server 127.127.1.0
fudge 127.127.1.0 stratum 10
NTPconf
fi
service ntp restart
## 7：设置Iscsi
apt-get install -y tgt ssh w3m unzip wget curl expect
## 安装iscsi客户端、安装rabbitmq
apt-get install -y open-iscsi open-iscsi-utils
# iscsitarget iscsitarget-dkms
# sed -i -e 's/false/true/' /etc/default/iscsitarget
# service iscsitarget restart
apt-get install -y rabbitmq-server memcached python-memcache
apt-get install -y kvm libvirt-bin qemu qemu-kvm
## 二：安装mysql和创建相关数据库
## Openstack的组件：nova，keystone，glance，都需要数据库。不过目前官方建议keystone，采用sqlite存储，而不用mysql存放。所以我们只需要创建nova和glance两个数据库就可以。
## 1：安装mysql
## 过程中，会提示你输入root密码。通过debconf-set-selections读取跳过这一步.
#MYSQL_PASSWD=${MYSQL_PASSWD:-"cloud1234"}
## apt-get install debconf debconf-utils
cat <<MYSQL_PRESEED | debconf-set-selections
mysql-server-5.1 mysql-server/root_password password $MYSQL_PASSWD
mysql-server-5.1 mysql-server/root_password_again password $MYSQL_PASSWD
mysql-server-5.1 mysql-server/start_on_boot boolean true
MYSQL_PRESEED
apt-get install --no-install-recommends -y mysql-server python-mysqldb
## chkconfig mysql on
## 让mysql支持外部访问
sed -i '/^bind-address/s/127.0.0.1/0.0.0.0/g' /etc/mysql/my.cnf 
service mysql restart
if [ ! -s /etc/apache2/httpd.conf ]; then
        echo "ServerName localhost" >> /etc/apache2/httpd.conf
        /etc/init.d/apache2 restart
fi
## 2：安装phpmyadmin （可选）
cat <<PHPmyadmin | debconf-set-selections
phpmyadmin phpmyadmin/reconfigure-webserver  text     apache2
phpmyadmin phpmyadmin/dbconfig-install       boolean  true
phpmyadmin phpmyadmin/app-password-confirm   password $MYSQL_PASSWD
phpmyadmin phpmyadmin/mysql/admin-pass       password $MYSQL_PASSWD
phpmyadmin phpmyadmin/password-confirm       password $MYSQL_PASSWD
phpmyadmin phpmyadmin/setup-password         password $MYSQL_PASSWD
phpmyadmin phpmyadmin/mysql/app-pass         password $MYSQL_PASSWD
PHPmyadmin
apt-get --no-install-recommends -yq install phpmyadmin
## 3：创建数据库
## nova数据库，   管理员：novadbadmin，密码是：cloud1234
## glance数据库，管理员：glancedbadmin，密码是：cloud1234
## 如果你修改密码，后面很多配置都需要相应更改。
##   mysql -uroot -p
##   CREATE DATABASE nova;
##   GRANT ALL PRIVILEGES ON nova.* TO 'novadbadmin'@'%' IDENTIFIED BY 'cloud1234';
##   CREATE DATABASE glance;
##   GRANT ALL PRIVILEGES ON glance.* TO 'glancedbadmin'@'%' IDENTIFIED BY 'cloud1234';
##   quit
mysql -uroot -p$MYSQL_PASSWD -e "DROP DATABASE IF EXISTS nova;"
mysql -uroot -p$MYSQL_PASSWD -e "CREATE DATABASE nova;"
mysql -uroot -p$MYSQL_PASSWD -e "GRANT ALL ON nova.* TO '$NOVA_DB_USERNAME'@'%' IDENTIFIED BY '$NOVA_DB_PASSWD';"
mysql -uroot -p$MYSQL_PASSWD -e "DROP DATABASE IF EXISTS glance;"
mysql -uroot -p$MYSQL_PASSWD -e "CREATE DATABASE glance;"
mysql -uroot -p$MYSQL_PASSWD -e "GRANT ALL ON glance.* TO '$GLANCE_DB_USERNAME'@'%' IDENTIFIED BY '$GLANCE_DB_PASSWD';"
if [ -e /var/lib/keystone/keystone.db ]; then
rm -rf /var/lib/keystone/keystone.db
fi
## 三：安装和配置keystone
## Openstack的各个组件，keystone是最难配置。搞定keystone，后面应该就没啥麻烦。
## 1：安装keystone
apt-get install -y keystone python-keystone python-keystoneclient
## 2：配置keystone
## 如果更改token，需要修改 /etc/keystone/keystone.conf 两个地方
##    默认定义的token就是ADMIN，web登录admin就是这个密码
##    [DEFAULT]
##    bind_host = 0.0.0.0
##    public_port = 5000
##    admin_port = 35357
##    admin_token = ADMIN
##    另外一个地方是
##    [catalog]
##    #driver = keystone.catalog.backends.sql.Catalog
##    driver = keystone.catalog.backends.templated.TemplatedCatalog
##    template_file = /etc/keystone/default_catalog.templates
##    
sed -i -e 's/keystone.catalog.backends.sql.Catalog/keystone.catalog.backends.templated.TemplatedCatalog\ntemplate_file = \/etc\/keystone\/default_catalog.templates/g' /etc/keystone/keystone.conf
sed -i -e "s/ADMIN/$ADMIN_TOKEN/g" /etc/keystone/keystone.conf
service keystone restart
## 3: 导入数据
## 这个比较有技术含量。通过修改devstack的keystone_data.sh 脚本。实现导入数据。如果你上面的设置都和我一样，那么其实直接运行这个脚本就可以。
## wget http://www.hastexo.com/system/files/user/4/keystone_data.sh_.txt -O keystone_data.sh
wget http://onestack.googlecode.com/files/gen_keystone_data.sh -O gen_keystone_data.sh
chmod +x gen_keystone_data.sh
## 运行脚本, 如果你修改的默认的用户名和密码，你需要修改脚本。修改两个地方
## 第一个是登录dashboard的admin的密码
## 第二个就是keystone的token
## #ADMIN_PASSWORD=${ADMIN_PASSWORD:-hastexo}
## ADMIN_PASSWORD=${ADMIN_PASSWORD:-admin}
## SERVICE_PASSWORD=${SERVICE_PASSWORD:-$ADMIN_PASSWORD}
## #export SERVICE_TOKEN="hastexo"
## export SERVICE_TOKEN="Centrin"
## export SERVICE_ENDPOINT="http://localhost:35357/v2.0"
## SERVICE_TENANT_NAME=${SERVICE_TENANT_NAME:-service}
sed -i -e "s/hastexo/$ADMIN_TOKEN/g" gen_keystone_data.sh
   
./gen_keystone_data.sh
##  顺利运行，会什么都没有输出
##  #echo $?
##  下面会输出0 ，表示正常。你就别再运行脚本
##  keystone --tenant=admin --username=admin --password=hastexo  --auth_url=http://127.0.0.1:5000/v2.0 user-list
##  看到这些，就说明keystone安装正常。
cat <<ENV_AUTH >> /etc/profile
export OS_TENANT_NAME=admin
export OS_USERNAME=admin
export OS_PASSWORD=ADMIN
export OS_AUTH_URL="http://localhost:5000/v2.0/"
ENV_AUTH
sed -i -e "s/ADMIN/$ADMIN_TOKEN/g" /etc/profile
export OS_TENANT_NAME=admin
export OS_USERNAME=admin
export OS_PASSWORD=$ADMIN_TOKEN
export OS_AUTH_URL="http://localhost:5000/v2.0/"
#source /etc/profile
## 检查检查
## # export | grep OS_
## declare -x OS_AUTH_URL="http://localhost:5000/v2.0/"
## declare -x OS_PASSWORD="hastexo"
## declare -x OS_TENANT_NAME="admin"
## declare -x OS_USERNAME="admin"
## 这个时候，直接运行
## root@node6:~# keystone user-list
## 四：安装和配置glance
## 1：安装软件
apt-get install -y glance glance-api glance-client glance-common glance-registry python-glance
## 2：配置/etc/glance/glance-api-paste.ini
## 修改文件最后3行，这些设置都是keystone导入数据的时候设置的。
sed -i -e "
       s/%SERVICE_TENANT_NAME%/admin/g;
       s/%SERVICE_USER%/admin/g;
       s/%SERVICE_PASSWORD%/$ADMIN_TOKEN/g;
    " /etc/glance/glance-api-paste.ini
   
## 3：设置 /etc/glance/glance-registry-paste.ini
sed -i -e "
       s/%SERVICE_TENANT_NAME%/admin/g;
       s/%SERVICE_USER%/admin/g;
       s/%SERVICE_PASSWORD%/$ADMIN_TOKEN/g;
    " /etc/glance/glance-registry-paste.ini
## 4：配置/etc/glance/glance-registry.conf
## 修改
## #sql_connection = sqlite:////var/lib/glance/glance.sqlite
## sql_connection = mysql://glancedbadmin:[email protected]/glance
## 在末尾添加两行
## [paste_deploy]
## flavor = keystone
GLANCE_API_CONF=${GLANCE_API_CONF:-"/etc/glance/glance-api.conf"}
GLANCE_REGISTRY_CONF=${GLANCE_REGISTRY_CONF:-"/etc/glance/glance-registry.conf"}
PUBLIC_IP=${PUBLIC_IP:-"192.168.139.50"}
sed -i '/sql_connection = .*/{s|sqlite:///.*|mysql://'"$GLANCE_DB_USERNAME"':'"$GLANCE_DB_PASSWD"'@'"$PUBLIC_IP"'/glance|g}' $GLANCE_API_CONF
cat <<EOF >>$GLANCE_API_CONF
[paste_deploy]
flavor = keystone
EOF
sed -i '/sql_connection = .*/{s|sqlite:///.*|mysql://'"$GLANCE_DB_USERNAME"':'"$GLANCE_DB_PASSWD"'@'"$PUBLIC_IP"'/glance|g}' $GLANCE_REGISTRY_CONF
cat <<EOF >>$GLANCE_REGISTRY_CONF
[paste_deploy]
flavor = keystone
EOF
## 5：配置/etc/glance/glance-api.conf
## 6：同步数据库
glance-manage version_control 0
glance-manage db_sync         
service glance-api restart && service glance-registry restart
## 7：验证glance服务是否正常
# glance index
## 没有任何的输出。表示正常。
## #echo $?
## 下面会输出0 ，表示正常。
## # glance --version
## glance 2012.1
## 五：安装配置nova
## 1:安装nova相关组件
apt-get install -y nova-api nova-cert nova-common nova-compute nova-compute-kvm nova-doc nova-network nova-objectstore nova-scheduler  nova-volume python-nova python-novaclient  nova-consoleauth python-novnc novnc
## 2：配置 /etc/nova/nova.conf
## 如果你是在虚拟机里测试Openstack。你需要把默认的虚拟化引擎从kvm改成qemu。
cat <<NOVAconf > /etc/nova/nova.conf
--dhcpbridge_flagfile=/etc/nova/nova.conf
--dhcpbridge=/usr/bin/nova-dhcpbridge
--logdir=/var/log/nova
--state_path=/var/lib/nova
--lock_path=/var/lock/nova
--allow_admin_api=true
--use_deprecated_auth=false
--auth_strategy=keystone
--scheduler_driver=nova.scheduler.simple.SimpleScheduler
--s3_host=192.168.139.50
--ec2_host=192.168.139.50
--rabbit_host=192.168.139.50
--cc_host=192.168.139.50
--nova_url=http://192.168.139.50:8774/v1.1/
--routing_source_ip=192.168.139.50
--glance_api_servers=192.168.139.50:9292
--image_service=nova.image.glance.GlanceImageService
--iscsi_ip_prefix=10.0.0
--sql_connection=mysql://novadbadmin:[email protected]/nova
--ec2_url=http://192.168.139.50:8773/services/Cloud
--keystone_ec2_url=http://192.168.139.50:5000/v2.0/ec2tokens
--api_paste_config=/etc/nova/api-paste.ini
--libvirt_type=kvm
--libvirt_use_virtio_for_bridges=true
--start_guests_on_host_boot=true
--resume_guests_state_on_host_boot=true
   
#novnc
--novnc_enabled=true
--novncproxy_base_url= http://192.168.139.50:6080/vnc_auto.html
--vncserver_proxyclient_address=127.0.0.1
--vncserver_listen=127.0.0.1
# network specific settings
--network_manager=nova.network.manager.FlatDHCPManager
--public_interface=eth0
--flat_interface=eth1
--flat_network_bridge=br100
--fixed_range=10.0.0.1/27
--floating_range=192.168.139.225/27
--network_size=32
--flat_network_dhcp_start=10.0.0.40
--flat_injected=False
--force_dhcp_release
--iscsi_helper=tgtadm
--connection_type=libvirt
--root_helper=sudo nova-rootwrap
#--verbose
--verbose=False
NOVAconf
## if ! kvm-ok 1>/dev/null 2>&1; then
##      sed -i -e "s/kvm/qemu/" /etc/nova/nova-compute.conf
## fi
sed -i -e "s/novadbadmin/$NOVA_DB_USERNAME/g;s/cloud1234/$NOVA_DB_PASSWD/g" /etc/nova/nova.conf
sed -i -e "s/192.168.139.50/$OUT_IP/g;s/192.168.139.225/$FLOAT_IP/g;" /etc/nova/nova.conf
sed -i -e "s/10.0.0.1/$IN_IP/g;s/10.0.0.40/$FLAT_IP/g;s/10.0.0/$IN_IP_PRE/g;" /etc/nova/nova.conf
## kvm or qemu?
sed -i -e "s/kvm/$VIRT_TYPE/g" /etc/nova/nova.conf
sed -i -e "s/kvm/$VIRT_TYPE/g" /etc/nova/nova-compute.conf
## 3：配置/etc/nova/api-paste.ini
sed -i -e "
       s/%SERVICE_TENANT_NAME%/admin/g;
       s/%SERVICE_USER%/admin/g;
       s/%SERVICE_PASSWORD%/$ADMIN_TOKEN/g;
    " /etc/nova/api-paste.ini
# 4：停止和重启nova相关服务
## 更改卷组，否则启动nova-volume会出错。
vgrename `hostname` nova-volumes
## 设置ipv4转发，否则外面能连接虚拟机，虚拟机访问不了外面
sysctl -w net.ipv4.ip_forward=1
##or:
##echo 1 > /proc/sys/net/ipv4/ip_forward
for a in libvirt-bin nova-network nova-compute nova-api nova-objectstore nova-scheduler novnc nova-volume nova-consoleauth; do service "$a" restart; done
## 六：安装和配置Dashbaord
## 1：安装dashbaord
apt-get install -y libapache2-mod-wsgi openstack-dashboard
/etc/init.d/apache2 restart
## 这个时候，你就可以登录dashboard
## http://192.168.139.50
## user:admin
## pass:ADMIN
## 之后通过前端web管理