Setting Up a Web UI Automation Testing Environment on CentOS with Selenium, Chrome, and VNC

Web UI automation testing often suffers from high maintenance overhead, but by applying proper configuration, isolation of locating and operating steps, and a well‑designed environment, the workload can be reduced.

Install GNOME desktop on the CentOS 7.1 server:

yum groupinstall "GNOME Desktop"

Install and configure VNC server to provide remote graphical access:

yum install tigervnc-server

Copy the service template and edit it:

/etc/systemd/system/vncserver@:1.service

[Unit] Description=Remote desktop service (VNC) After=syslog.target network.target [Service] Type=forking User=autotestuser ExecStart=/usr/bin/vncserver :1 -geometry 1280x1024 -depth 16 -securitytypes=none -fp /usr/share/X11/fonts/misc ExecStop=/usr/bin/vncserver -kill :1 [Install] WantedBy=multi-user.target

Enable and start the VNC service, then verify it is listening:

systemctl enable vncserver@:1.service systemctl start vncserver@:1.service netstat -lntp | grep vnc

Install the stable Google Chrome browser (ensure the repository is configured):

yum install google-chrome-stable

Install Python 2 pip and Selenium:

yum install python2-pip pip install -U selenium

Download the matching chromedriver from chromedriver site and place it in PATH or specify its location in scripts, making sure the driver version matches the Chrome version.

For large test suites, use Python's multiprocessing module to run tests in parallel across multiple browser instances, each in its own process, eliminating data dependencies and significantly shortening regression cycles.

By following these steps, a stable, scalable web UI automation environment is established, enabling efficient test execution and maintenance.