Im testing some selenium test's in Jenkins and they are crashing, I read in a stackoverflow post that it is because Jenkins uses a virtual machine and it can be a memory problem, the solution is to use somethink like this: chrome_options.add_argument('--disable-dev-shm-usage') but the problem is that I dont know where I should put this code in order to solve the problem, I dont know if I use the code on docker.compose or in the tests-selenium.side.
chrome_options.add_argument('--disable-dev-shm-usage')
This is a chrome option, add this (tests-selenium.side) where you have Options instance.
You must have
options = webdriver.ChromeOptions()
written in your code somewhere.
Then redeploy your script in Jenkins VM and it should work.
Related
I am using Selenium to open a web site, login and copy some information from one web site. However it is happening on my work station and have a display monitor.
my IT team wants to move this process to a virtual server which does not have a monitor.
1.Will this work - even if we install Chrome of Firefox on the server
2. Can we Chrome - headless to make this happen
3. Any other way - we can think of using Xserver
Please let me know.
No . To run your script you don't need to have monitor. You can access your virtual machine through remote connection and you can start the execution from that machine. Once the execution started, you can close the remote desktop session and execution will continue to run on remote machine or virtual server.
I hope this helps. Please let me know if you have any further questions.
1.Will this work - even if we install Chrome or Firefox on the server - Yes it will work
2.Can we Chrome - headless to make this happen - If you are going to use virtual server just for execution,then you don't need to run in headless mode. Headless execution is needed for environments where you don't need a visible UI shell. Below code will help you run your script in headless mode
from selenium import webdriver
from selenium.webdriver.chrome.options import Options
options = Options()
options.add_argument('--headless')
options.add_argument('--disable-gpu') # Last I checked this was necessary.
driver = webdriver.Chrome("/usr/local/bin/chromedriver", chrome_options=options)
driver.get("https://google.com")
#code to extract the details
driver.quit()
3.Any other way - we can think of using Xserver - Not sure
Chrome headless should solve your problem here -- I've done this in the past with some of my automation and had success.
Just remember to use ChromeOptions to add the --headless flag. You may need to tweak some other ChromeOptions as well -- I also had to add --disable-gpu and --window-size=1920,1200 to get mine working just right.
I can handle it with FF web driver (selenium python) but when I change phantomjs the driver .it didnt handle it. is it possible to handle it?
the program must work on server so what must I do?
In short, no, this is not currently possible. The PhantomJS ghostdriver does not implement the primitives for handling alerts/prompts. See https://github.com/detro/ghostdriver/issues/20
This is also unlikely to change since development of PhantomJS has been discontinued.
the program must work on server so what must I do?
Your best option is to use the headless versions of Chrome or Firefox, which can run in a headless environment like PhantomJS just fine. Both Chrome (chromedriver) and Firefox (geckodriver) implement the necessary primitives for handling alerts/prompts.
Another option is that you can use a virtual screen program (e.g. xvfb) to enable you to use a headed browser in a headless environment.
I have a Java Application, which controls an automated GUI test in a FF-Browser via Selenium WebDriver Libraray. The Java App reads test cases from a database and executes them according to the code logic.
For instance, if the app reads in a Field, it'll search it by using the "findElement"-method from the Selenium framework. I do not use any test scripts for Selenium.
Currently this is happening on a local workingstation of an employee.
Now I want to move this whole environment into a Docker container.
Is it even possible to instantiate a Firefox Browser in a Container?
btw: I do not need to see the actual GUI of my browser.
And secondly:
There are several containers with selenium on dockerhub ready to use, but these do not fit my surroundings am I right?
As far as I know the SeleniumGrid expects testscripts and cannot be executed through runtime.
I open up a Linux VM (Debian:Jessie distribution) with Vagrant, in which then runs Docker.
I am still a beginner with Docker.
I couldn't find any question around here regarding my purpose.
Thanks in advance!
Is it even possible to instantiate a Firefox Browser in a Container?
Yes. The simplest way to do this is would be using the selenium images on Docker Hub.
There are several containers with selenium on dockerhub ready to use, but these do not fit my surroundings am I right?
If you think the Selenium images don't work for you because they are all based on Selenium Grid, you can use the StandaloneFirefox and StandaloneChrome images instead. These are individual instances, they do not use Selenium Grid.
BTW, the non-Debug Selenium images do not have a GUI. You mentioned you didn't need to see the browsers running so these should be fine. If you do need to see the browsers, the Debug images have a VNC server installed so you can run the image, connect with a VNC client, and watch the browsers run the tests.
I am doing a lot of testing with the Chrome Web Driver within Selenium. The problem is that every time I run it, it has to re-download all the site images which takes time. It does not keep these images cached.
So I would like to set which files to render and which not too. Obviously I want Javascript and CSS files to still download. But I particularily want to turn off images.
Is this possible? If not, is there a way to enable caching? So the next time I run the progran it can get the images from the local cache.
The solution is to load the same chrome profile again, it should (may not) ensure that images & other similar things are cached.
Here is how to load a particular profile :
https://code.google.com/p/selenium/wiki/ChromeDriver
DesiredCapabilities capabilities = DesiredCapabilities.chrome();
capabilities.setCapability("chrome.switches", Arrays.asList("--user-data-dir=/path/to/profile/directory"));
WebDriver driver = new ChromeDriver(capabilities);
A similar search on SO, gave this result - Load Chrome Profile using Selenium WebDriver using java you might bother to take a look.
Caching from one session to another is not possible as far as I know.
However, it is possible to remove rendering the page if you run it headless. This will load the page, but not render it (make it visible, load images).
You can use HTMLUnitDriver, which is the standard, but somewhat outdated, or you can use PhantomDriver, which has a more modern version of Javascript.
"You can use HTMLUnitDriver" >>> careful with this!
Please note that since HTMLUnitDriver is not the same implementation as the real browser, you may or may not get the same results in both.
So if you start hitting weird issues, like running the test with Chrome passes but running the test with HTMLUnitDriver does not, consider just running your tests with the browser driver. I've heard of looong troubleshooting stories from colleagues and they had to give up on running their suites in headless mode (ie, using HTMLUnitDriver).
NOTE: since I cannot leave a comment on the selected answer, I am "forced" to leave it as an answer. If somebody can help me convert it in a comment, I will appreciate! thanks
For caching images you would need something like squid or haproxy, and then proxy your selenium browser through it.
I am looking for a tool to functionally test my Yii/ Node.js web application. The first thing I looked into was Selenium. The app runs on a headless Ubuntu server so successfully setting up xvfb and run a test was really painfull and drove me to another tool.
The error I kept getting is:
Xlib: extension "RANDR" missing on display :0
The other tool was Casperjs along with Phantomjs . Aside from the 5 minute setting up, I wrote few tests and integrated all with Jenkins CI. I really believe there should be more tools like this one. So I feel I've earned something on the short term, but I'm afraid that on the long term I'll hit a dead end. Could you give me some feedback? Am I going the wrong road?
Another thing that's crossing my mind is to setup the Selenium RC and Jenkins on a Windows machine with all browsers set up. I think this will give my tests a better and more accurate perspective.
* I would also like to be able to do some parallel functional tests (interactions) since the website is socket-driven. Does Selenium handle that?
First off, don't use Selenium RC if you can avoid it, it's officially deprecated in favor of Selenium Webdriver (also known as Selenium 2).
About headlessness - Webdriver can be easily run on top of HtmlUnit and also PhantomJS - both should work right away.
If you want to run the tests on actual browsers, people have been successful in running that on headless systems, too, but as you said, it's a pain. Instead, you can run the tests remotely on a different machines with your headless server commanding them all.