附加安装选项¶

安装Usage服务（可选）¶

管理服务器配置完毕后，你可以选择安装Usage服务器。Usage服务器从系统事件中提取数据，便于对账户进行使用计费。

当存在多台管理服务器时，可以选择在它们上面安装任意数量的Usage服务器。Usage服务器会协调处理。考虑到高可用性，所以最少在两台管理服务器中安装Usage服务器。

SSL (可选)¶

CloudStack默认提供HTTP的访问方式。有许多的技术和网站选择使用SSL。因此，我们先抛下CloudStack所使用的HTTP，假设站点使用SSL是一个典型实践。

CloudStack使用Tomcat作为服务容器。由于CloudStack网站中断SSL会话，可能会开启Tomcatde的SSL访问。Tomcat的SSL配置描述请查阅： http://tomcat.apache.org/tomcat-6.0-doc/ssl-howto.html.

数据库复制（可选）¶

CloudStack支持MySQL节点间的数据库复制。这个通过标准的MySQL复制功能来实现。你这样做可能希望防止MySQL服务器或者存储损坏。MySQL复制使用master/slave的模型。master节点是直接为管理服务器所使用。slave节点为备用，接收来自master节点的所有写操作，并将它应用于本地冗余数据库。以下是实施数据库复制的步骤。

1. 确保这是一个全新安装且没有数据的master数据库节点。

2. 编辑master数据库的my.cnf，在[mysqld]的datadir下增加如下部分。

   log_bin=mysql-bin
   server_id=1
   

   考虑到其他的服务器，服务器id必须是唯一的。推荐的方式是将master的ID设置为1，后续的每个slave节点序号大于1，使得所有服务器编号如：1，2，3等。

3. 重启MySQL服务。如果是RHEL/CentOS系统，命令为：

   # service mysqld restart
   

   如果是Debian/Ubuntu系统，命令为：

   # service mysql restart
   

4. 在master上创建一个用于复制的账户，并赋予权限。我们创建用户”cloud-repl”，密码为”password”。假设master和slave都运行在172.16.1.0/24网段。

   # mysql -u root
   mysql> create user 'cloud-repl'@'172.16.1.%' identified by 'password';
   mysql> grant replication slave on *.* TO 'cloud-repl'@'172.16.1.%';
   mysql> flush privileges;
   mysql> flush tables with read lock;
   

5. 离开当前正在运行的MySQL会话。

6. 在新的shell中打开第二个MySQL会话。

7. 检索当前数据库的位置点。

   # mysql -u root
   mysql> show master status;
   +------------------+----------+--------------+------------------+
   | File             | Position | Binlog_Do_DB | Binlog_Ignore_DB |
   +------------------+----------+--------------+------------------+
   | mysql-bin.000001 |      412 |              |                  |
   +------------------+----------+--------------+------------------+
   

8. 注意你数据库实例所返回的文件及位置点。

9. 退出该会话。

10. 完成master安装。返回到master的第一个会话，取消锁定并退出MySQL。

    mysql> unlock tables;
    

11. 安装并配置slave节点。在slave服务器上，运行如下命令。

    # yum install mysql-server
    # chkconfig mysqld on
    

12. 编辑my.cnf，在[mysqld]的datadir下增加如下部分。

    server_id=2
    innodb_rollback_on_timeout=1
    innodb_lock_wait_timeout=600
    

13. 重启MySQL。对于RHEL/CentOS系统，使用”mysqld”

    # service mysqld restart
    

    对于Ubuntu/Debian系统，使用”mysql.”

    # service mysql restart
    

14. 引导slave连接master并进行复制。使用上面步骤中得到数据来替换IP地址，密码，日志文件，以及位置点。

    mysql> change master to
        -> master_host='172.16.1.217',
        -> master_user='cloud-repl',
        -> master_password='password',
        -> master_log_file='mysql-bin.000001',
        -> master_log_pos=412;
    

15. 在slave上启动复制功能。

    mysql> start slave;
    

16. 在slave上可能需要开启3306端口，就像之前在master上所做的一样。

    这对复制来说不是必须的。但如果没有做，当需要进行数据库切换时，你仍然需要去做。

Amazon Web Services 接口¶

http://localhost:8096/client/api?command=updateConfiguration&name=enable.ec2.api&value=true
http://localhost:8096/client/api?command=updateConfiguration&name=enable.ec2.api&value=true

$ export EC2_CERT=/path/to/cert.pem
$ export EC2_PRIVATE_KEY=/path/to/private_key.pem
$ export EC2_URL=http://localhost:7080/awsapi
$ export EC2_HOME=/path/to/EC2_tools_directory

--connection-timeout TIMEOUT

--request-timeout TIMEOUT

ec2-run-instances 2 –z us-test1 –n 1-3 --connection-timeout 120 --request-timeout 120

#!/usr/bin/env python

import sys
import os
import boto
import boto.ec2

region = boto.ec2.regioninfo.RegionInfo(name="ROOT",endpoint="localhost")
apikey='GwNnpUPrO6KgIdZu01z_ZhhZnKjtSdRwuYd4DvpzvFpyxGMvrzno2q05MB0ViBoFYtdqKd'
secretkey='t4eXLEYWw7chBhDlaKf38adCMSHx_wlds6JfSx3z9fSpSOm0AbP9Moj0oGIzy2LSC8iw'

def main():
    '''Establish connection to EC2 cloud'''
    conn = boto.connect_ec2(aws_access_key_id=apikey,
                            aws_secret_access_key=secretkey,
                            is_secure=False,
                            region=region,
                            port=7080,
                            path="/awsapi",
                            api_version="2010-11-15")

    '''Get list of images that I own'''
    images = conn.get_all_images()
    print images
    myimage = images[0]
    '''Pick an instance type'''
    vm_type='m1.small'
    reservation = myimage.run(instance_type=vm_type,security_groups=['default'])

if __name__ == '__main__':
     main()

#!/usr/bin/env python

import sys
import os
from boto.s3.key import Key
from boto.s3.connection import S3Connection
from boto.s3.connection import OrdinaryCallingFormat

apikey='ChOw-pwdcCFy6fpeyv6kUaR0NnhzmG3tE7HLN2z3OB_s-ogF5HjZtN4rnzKnq2UjtnHeg_yLA5gOw'
secretkey='IMY8R7CJQiSGFk4cHwfXXN3DUFXz07cCiU80eM3MCmfLs7kusgyOfm0g9qzXRXhoAPCH-IRxXc3w'

cf=OrdinaryCallingFormat()

def main():
    '''Establish connection to S3 service'''
    conn = S3Connection(aws_access_key_id=apikey,aws_secret_access_key=secretkey, \
                        is_secure=False, \
                        host='localhost', \
                        port=7080, \
                        calling_format=cf, \
                        path="/awsapi/rest/AmazonS3")

    try:
        bucket=conn.create_bucket('cloudstack')
        k = Key(bucket)
        k.key = 'test'
        try:
            k.set_contents_from_filename('/Users/runseb/Desktop/s3cs.py')
        except:
            print 'could not write file'
            pass
    except:
        bucket = conn.get_bucket('cloudstack')
        k = Key(bucket)
        k.key = 'test'
        try:
            k.get_contents_to_filename('/Users/runseb/Desktop/foobar')
        except:
            print 'Could not get file'
            pass

    try:
        bucket1=conn.create_bucket('teststring')
        k=Key(bucket1)
        k.key('foobar')
        k.set_contents_from_string('This is my silly test')
    except:
        bucket1=conn.get_bucket('teststring')
        k = Key(bucket1)
        k.key='foobar'
        k.get_contents_as_string()

if __name__ == '__main__':
    main()

Adding new Guest Physical Network to existing Zone¶

In some cases it might be necessary to add additional Guest Physical Network to already existing Advanced Zone. One example might be that we have been running Zone with Guest Physical Network that uses VXLAN as isolation method (i.e. KVM traffic label set to “bond.150), and we also need to start using Private Gateway functionality, which would not work with current Guest Physical Network Traffic Label (for explanation why this would not work, please see “adding-priv-gw-vpc”.

So in example below, we would like to create new Guest Physical Network (Traffic Label set to “bond0” in this example) and later use this Guest Physical Network for other purposes (to provision Private Gateways on top of it)

Depending on the CloudStack version, we can probably accomplish this only by API commands (i.e. CloudMonkey) and direct database changes.

First obtain Zone ID, and then create Physical Network command:

cloudmonkey> list zones filter=id | grep id | head -n1
cloudmonkey> create physicalnetwork name="bond0 Private Gateways" broadcastdomainrange=zone zoneid=xxxxxx isolationmethod=VLAN

where “xxxxxxx” is the proper zone ID (output from the first command above).

Here we created new Physical Network (not yet of any type…) inside given zone.

Please note that new Physical Network can not be added, while Zone is in Enabled state!

It is safe to disable Zone (for few seconds) which will just prevent users from provisioning new resources in this Zone, then create Physical Network per instructions from above, and then enable the Zone (this can be done via GUI or CloudMonkey).

Since Physical Network is in Disabled state by default, we need to enable it by:

cloudmonkey> list physicalnetworks name="bond0 Private Gateways" | grep id | head -n1
cloudmonkey> update physicalnetwork state=Enabled id=xxxxx

where xxxxx is the correct ID of Physical Network (output from the first command above).

After Physical Network is created, although we selected VLAN as isolation method, it may happen (depending on CloudStack version) that there will be no DB record created (which defines VLAN as isolation method for this new network), so we need to check this and fix it:

cloudmonkey> list physicalnetworks name="bond0 Private Gateways" | grep isolationmethods

If above commands gives no output, that means that VLAN was not set as Isolation Method inside DB, in cloud.physical_network_isolation_methods table, so we need to set it:

mysql> select id from cloud.physical_network where name="bond0 Private Gateways"
mysql> select max(id) from cloud.physical_network_isolation_methods;

Write down the id and max(id) values (i.e. 204, 10)

mysql> INSERT INTO cloud.physical_network_isolation_methods (id, physical_network_id, isolation_method) VALUES ('11', '204', 'VLAN');

where 11 is actually the next number after the max(id) value, and 204 is the ID of the Physical Network (check the image below)

Now, that we have set correct Isolation Method for the new Physical Network, and enabled it, we also need to do most critical part, to define that this Physical Network caries Guest traffic. This action is also done via database change.

We want to effectively clone the existing Guest network row from cloud.physical_network_traffic_types table, and then change needed values: ID, UUID, PHYSICAL_NETWORK_ID and KVM_NETWORK_LABEL (in our case, we are using KVM, but same goes for other HyperVisor types)

After this last step, we are ready to tag these 2 Guest Physical Networks and Network Offerings (regular ones for VPC Guest networks, and the special one used for Private Gateways) and then use these two Guest Networks as originally planned.

For more details on Network tagging, please see “adding-priv-gw-vpc”.