自动化测试时,我们有很多用例,比如2000条用例,按照顺序执行,每条用例执行1分钟,那需要2000分钟;
什么概念?2000分钟就30多个小时,如果是冒烟测试,估计还没人工跑的快;
还有,如果是线上发布,跑完2000条用例就太浪费时间了;
那如果我们让我们让用例分布式执行,是不是可以节省很多时间?
如果数据量很大,我们使用1台压测机,可能并发压力过大;
那就需要选择使用多台压测机(比如Jmeter的 Agent/负载机);
这样也是一种分布式压测或者分布式性能测试场景。
所以总结来说,其实就是为了提升效率和质量。
特点 | 说明 |
---|---|
网格化 | 多节点互联互通,可资源共享 |
分布性 | 地域和计算机上,协同工作、负载均衡、可扩展性、高可用性 |
开放性 | 可移植性、可互操作性、可伸缩性、易获得性 |
实时性 | 各种信息都必须是实时的 |
动态性 | 测试过程对象和活动动态映射 |
处理不确定性 | 具有处理不确定性的能力 |
容错及安全性 | 容错能力强,可靠性高、安全性好 |
技术点 | 要求 |
---|---|
分布式环境 | 获取全局状态,能够方便地监视和操纵测试过程;集中式的分布式策略。 |
分布式环境下的节点通信 | 稳定的通信环境;适合用基于消息通信的方式来实现。 |
测试任务调度 | 静态调度、动态调度和混合调度。 |
用例之间是独立且没有依赖关系,完全独立运行;
用例执行没有顺序,随机顺序都能正常执行;
每个用例都能重复运行,运行结果不会影响其他用例。
pytest-xdist
让自动化测试用例分布式执行,节省测试时间,属于进程级别的并发;
使用以下方法安装:
pip3 install pytest-xdist
C:\Users\Administrator>pip3 install pytest-xdist
Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple
Requirement already satisfied: pytest-xdist in d:\python37\lib\site-packages (1.31.0)
Requirement already satisfied: six in d:\python37\lib\site-packages (from pytest-xdist) (1.15.0)
Requirement already satisfied: execnet>=1.1in d:\python37\lib\site-packages (from pytest-xdist) (1.8.0)
Requirement already satisfied: pytest>=4.4.0in d:\python37\lib\site-packages (from pytest-xdist) (6.2.4)
Requirement already satisfied: pytest-forked in d:\python37\lib\site-packages (from pytest-xdist) (1.1.3)
Requirement already satisfied: apipkg>=1.4in d:\python37\lib\site-packages (from execnet>=1.1->pytest-xdist) (1.5)
Requirement already satisfied: toml in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (0.10.2)
Requirement already satisfied: attrs>=19.2.0in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (20.3.0)
Requirement already satisfied: colorama in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (0.4.4)
Requirement already satisfied: atomicwrites>=1.0in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (1.4.0)
Requirement already satisfied: pluggy<1.0.0a1,>=0.12in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (0.13.1)
Requirement already satisfied: py>=1.8.2in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (1.10.0)
Requirement already satisfied: importlib-metadata>=0.12in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (2.1.1)
Requirement already satisfied: packaging in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (20.8)
Requirement already satisfied: iniconfig in d:\python37\lib\site-packages (from pytest>=4.4.0->pytest-xdist) (1.1.1)
Requirement already satisfied: zipp>=0.5in d:\python37\lib\site-packages (from importlib-metadata>=0.12->pytest>=4.4.0->pytest-xdist) (1.2.0)
Requirement already satisfied: pyparsing>=2.0.2in d:\python37\lib\site-packages (from packaging->pytest>=4.4.0->pytest-xdist) (2.4.7)
测试运行并行化;
在子进程中重复运行测试;
可指定不同的Python解释程序或不同的平台,并行运行测试。
import pytest
import time
class TestCase01():
def test_case_01(self):
time.sleep(1)
print("case01$$$$$$$$$$$$$$$$$$$$$")
def test_case_02(self):
time.sleep(1)
print("case02$$$$$$$$$$$$$$$$$$$$$")
def test_case_03(self):
time.sleep(1)
print("case03$$$$$$$$$$$$$$$$$$$$$")
def test_case_04(self):
time.sleep(1)
print("case04$$$$$$$$$$$$$$$$$$$$$")
def test_case_05(self):
time.sleep(1)
print("case05$$$$$$$$$$$$$$$$$$$$$")
def test_case_06(self):
time.sleep(1)
print("case06$$$$$$$$$$$$$$$$$$$$$")
class TestCase02():
def test_case_07(self):
time.sleep(1)
print("case07$$$$$$$$$$$$$$$$$$$$$")
def test_case_08(self):
time.sleep(1)
print("case08$$$$$$$$$$$$$$$$$$$$$")
def test_case_09(self):
time.sleep(1)
print("case08$$$$$$$$$$$$$$$$$$$$$")
if __name__ == '__main__':
pytest.main(["-s", "test_xdist.py"])
执行结果如下,使用了9.14s:
test_xdist.py::TestCase01::test_case_01
test_xdist.py::TestCase01::test_case_02
test_xdist.py::TestCase01::test_case_03
test_xdist.py::TestCase01::test_case_04
test_xdist.py::TestCase01::test_case_05
test_xdist.py::TestCase01::test_case_06
test_xdist.py::TestCase02::test_case_07 PASSED [ 11%]case01$$$$$$$$$$$$$$$$$$$$$
PASSED [ 22%]case02$$$$$$$$$$$$$$$$$$$$$
PASSED [ 33%]case03$$$$$$$$$$$$$$$$$$$$$
PASSED [ 44%]case04$$$$$$$$$$$$$$$$$$$$$
PASSED [ 55%]case05$$$$$$$$$$$$$$$$$$$$$
PASSED [ 66%]case06$$$$$$$$$$$$$$$$$$$$$
PASSED [ 77%]case07$$$$$$$$$$$$$$$$$$$$$
test_xdist.py::TestCase02::test_case_08 PASSED [ 88%]case08$$$$$$$$$$$$$$$$$$$$$
test_xdist.py::TestCase02::test_case_09 PASSED [100%]case08$$$$$$$$$$$$$$$$$$$$$
============================== 9 passed in9.14s ==============================
执行命令:
pytest -s -n auto test_xdist.py
结果如下,用时4.51s,可见分布式执行后大大缩短了测试时间:
(venv) F:\pytest_study\test_case\test_j>pytest -s -n auto test_xdist.py
============================================ test session starts =============================================
platform win32 -- Python 3.7.0, pytest-6.2.4, py-1.10.0, pluggy-0.13.1
rootdir: F:\pytest_study, configfile: pytest.ini
plugins: allure-pytest-2.8.12, assume-2.4.3, cov-2.8.1, forked-1.1.3, html-2.0.1, metadata-1.8.0, ordering-0.6,
repeat-0.9.1, rerunfailures-10.3, xdist-1.31.0
gw0 [9] / gw1 [9] / gw2 [9] / gw3 [9] / gw4 [9] / gw5 [9] / gw6 [9] / gw7 [9]
.........
============================================= 9 passed in4.51s ==============================================
-n auto
:可以自动检测到系统的CPU
核数;
使用auto
利用了所有CPU
来跑用例;
也可以指定使用几个CPU
来跑用例:
# x为cpu个数
pytest -s -n x
如下可以看到使用两个CPU
来跑用例时长为6.27s:
(venv) F:\pytest_study\test_case\test_j>pytest -s -n 2 test_xdist.py
============================================ test session starts =============================================
platform win32 -- Python 3.7.0, pytest-6.2.4, py-1.10.0, pluggy-0.13.1
rootdir: F:\pytest_study, configfile: pytest.ini
plugins: allure-pytest-2.8.12, assume-2.4.3, cov-2.8.1, forked-1.1.3, html-2.0.1, metadata-1.8.0, ordering-0.6,
repeat-0.9.1, rerunfailures-10.3, xdist-1.31.0
gw0 [9] / gw1 [9]
.........
============================================= 9 passed in6.27s ==============================================
命令如下:
pytest -s -n auto --html=report.html --self-contained-html
运行结果:
pytest -s -n auto test_xdist.py --html=report.thml --self-contained-htm
l
gw0 [9] / gw1 [9] / gw2 [9] / gw3 [9] / gw4 [9] / gw5 [9] / gw6 [9] / gw7 [9]
.........
------------------ generated html file: file://F:\pytest_study\test_case\test_j\report.thml ------------------
============================================= 9 passed in4.68s ==============================================
pytest-xdist
执行默认是无须的;
可通过?--dist
?参数来控制顺序;
参数 | 说明 |
---|---|
--dist=loadscope | 同一个模块module 下的函数和同一个测试类class 下的方法来分组 |
--dist=loadfile | 同一个文件名来分组 |
pytest-xdist
没有内置的支持来确保会话范围的夹具仅执行一次;
可使用FileLock
方法仅仅产生一次fixture
数据:
import pytest
from filelock import FileLock
@pytest.fixture(scope="session")
def login():
print("====登录===")
with FileLock("session.lock"):
name = "zhang"
password= "123456"
# web ui自动化
# 声明一个driver,再返回
# 接口自动化
# 发起一个登录请求,将token返回都可以这样写
yield name, password
print("====退出====")
最后感谢每一个认真阅读我文章的人,礼尚往来总是要有的,虽然不是什么很值钱的东西,如果你用得到的话可以直接拿走:
这些资料,对于【软件测试】的朋友来说应该是最全面最完整的备战仓库,这个仓库也陪伴上万个测试工程师们走过最艰难的路程,希望也能帮助到你!??