标签：
注意点：在测试nova，在配置文件里面如果不指定网络id，那么默认是外网的网络（该网络是共享的），如果想要指定网络，那么该网络必须是共享的状态，否则将会报错：无法发现网络。如果测试多于50台的虚拟机需要修改默认值，因为默认值是有限制的可用如下命令查看：
[root@rally nova]# nova absolute-limits +-------------------------+---------+ | Name | Value | +-------------------------+---------+ | maxServerMeta | 128 | | maxPersonality | 100 | | maxImageMeta | 128 | | maxPersonalitySize | 10240 | | maxTotalRAMSize | 5120000 | | maxSecurityGroupRules | 20 | | maxTotalKeypairs | 100 | | totalRAMUsed | 58960 | | maxSecurityGroups | 10 | | totalFloatingIpsUsed | 0 | | totalInstancesUsed | 24 | | totalSecurityGroupsUsed | 1 | | maxTotalFloatingIps | 10 | | maxTotalInstances | 200 | | totalCoresUsed | 51 | | maxTotalCores | 500 | +-------------------------+---------+
修改默认值可以在界面，管理员->默认值，进行修改；其次可以在运行nova-api服务的节点上修改配置文件nova.conf
安装rally：
[root@rally nova]#git clone https://github.com/stackforge/rally.git
[root@rally rally]# /usr/bin/easy_install pip （安装pip,有就不需要装了）
[root@rally rally]# mkdir /root/.pip
[root@rally rally]# vim /root/.pip/pip.conf （设置pip源，加快安装速度）
[global]
index-url=http://pypi.douban.com/simple
[root@rally nova]#pip install -U virtualenv 安装虚拟环境（可装可不装，一般为了系统的干净还是装个虚拟环境）
[root@rally nova]#virtualenv .venv
[root@rally nova]#. .venv/bin/activate
[root@rally nova]#yum install gcc-c++ （安装编译工具）
[root@rally nova]#cd rally
[root@rally rally]#./install_rally.sh
....
======================================================================
Information about your Rally installation:
* Method: system
* Database at: /var/lib/rally/database
* Configuration file at: /etc/rally
======================================================================
安装完成，数据库的配置这里暂时没做。
创建一个随便命名的.json格式文件：
[root@rally rally]# cat existing.json { "type": "ExistingCloud", "auth_url": ‘http://172.16.105.189:35357/v2.0/‘, "admin":{ "username": admin, "password": admin, "tenant_name": admin } }
[root@rally rally]# rally deployment create --file=existing.json --name=existing
+--------------------------------------+----------------------------+----------+------------------+--------+
| uuid | created_at | name | status | active |
+--------------------------------------+----------------------------+----------+------------------+--------+
| 15ef811a-6460-407c-b391-c0f435f1ae54 | 2015-03-05 10:09:36.468176 | existing | deploy->finished | |
+--------------------------------------+----------------------------+----------+------------------+--------+
Using deployment: 15ef811a-6460-407c-b391-c0f435f1ae54
[root@rally rally]# rally deployment check
keystone endpoints are valid and following services are available:
+------------+-----------+-----------+
| services | type | status |
+------------+-----------+-----------+
| ceilometer | metering | Available |
| cinder | volume | Available |
| cinderv2 | volumev2 | Available |
| glance | image | Available |
| keystone | identity | Available |
| neutron | network | Available |
| nova | compute | Available |
| nova_ec2 | ec2 | Available |
| novav3 | computev3 | Available |
+------------+-----------+-----------+
如此便可继续往下进行OpenStack基准测试了。
举个测试启动虚拟机的例子：boot.json文件内容如下：
doc/samples/tasks/scenarios/nova/boot.json
[root@rally nova]# cat boot.json { "NovaServers.boot_server": [ { "args": { "flavor": { "name": "m1.large" }, "image": { "name": "Windows_Server_2008R2_SP1_Standard_64bit" }, "nics":[{"net-id": "c7048568-c966-4d57-a927-90dd8830fb96"}],(默认是没有这行的) }, "runner": { "type": "constant", "times": 100, （测试次数100） "concurrency": 2 （并发数） }, "context": { "users": { "tenants": 3, （） "users_per_tenant": 2 } } } ] }
[root@rally nova]# . /root/rally/.venv/bin/activate
[root@rally nova]# rally -v task start boot.json（/root/rally/doc/samples/tasks/scenarios/nova这是当前的目录）
以上的图是测试结果。
相关链接：https://wiki.openstack.org/wiki/Rally
https://rally.readthedocs.org/en/latest/tutorial/step_0_installation.html
https://rally.readthedocs.org/en/latest/tutorial/step_1_setting_up_env_and_running_benchmark_from_samples.html
标签：
原文地址：http://www.cnblogs.com/wtfbk/p/4316335.html
