Hi.. I am using nightwatch CliRunner in a nodejs as a service. I am trying to enable the parallel testing by setting
test_workers: {
enabled: true,
workers: "auto"
},
After adding these two, I am getting the message "Started child process for: xxxx.spec.js", but it is failing soon afterwards with error "Address already in use ::: 8081" This address is actually the address of the nodejs server. So is it trying to run the server again?
Is there a way to run the tests in parallel when using this in nodejs server instead of cli tool?
I understand that nightwatch creates child processes and runs the node command again with some additional args. My requirement is I am running these tests as part of a nodejs server. If I enable the test_workers, the child processes are running the command to start the server and it is failing. I am using CLIRunner inside nodejs server to run the tests. Is there a way to run tests parallelly in this scenario using saucelabs? Can we run multiple CLIRunner instances passing single file to each of these instance? Can someone please help in figuring this out?
so what you're trying to do is to use Nightwatch programatically and use the parallel feature? That might not work, but I'd like to see a more complete example. Another way of handling it is to start your nodejs server from nightwatch, as part of the global hook.
Hi! Who can help with async in specs nightwatch? https://stackoverflow.com/questions/69451363/nighwatch-ignores-await-in-tests
Hi all, Nightwatch v2-alpha.3 is published in NPM. It has a tight integration with Cucumber. There is also a repository with examples, boilerplate, and a lot of predefined cucumber steps: https://github.com/gravityvi/cucumber-nightwatch-boilerplate
Hi all!
I have a test and fail message
buildPlanPage.assert.containsText('@activeButton','Gravity Only')
Testing if element <Element [name=@activeButton]> contains text 'Gravity Only' in 5000ms - expected "contains text 'Gravity Only'" but got: "does not contain 'Gravity Only'"
Is it possible somehow to have message like
expected "contains text 'Gravity Only'" but got: "real text that element has" ?
I have a test and fail message
buildPlanPage.assert.containsText('@activeButton','Gravity Only')
Testing if element <Element [name=@activeButton]> contains text 'Gravity Only' in 5000ms - expected "contains text 'Gravity Only'" but got: "does not contain 'Gravity Only'"
Is it possible somehow to have message like
expected "contains text 'Gravity Only'" but got: "real text that element has" ?
Hey guys. If I use the
--suiteRetries option, my github action runs forever (it sais passed successfully but does not exit). I run it using the following command:xvfb-run --auto-servernum --server-args='-screen 0, 1920x1080x24' yarn test -e local --suiteRetries 3 --verbose. Did any of you ever meet this problem?
And one more update. I downgraded nightwatch to v1.7 and all works good. It seems to be a bug in the v2.
1 reply
Hello - Thank you very much for nightwatchjs. Been using for a few months and enjoying. Just came to a requirement that a test simulate various network connectivity conditions, including no connectivity for X seconds. Familiar with how to do this with puppeteer, and webdriver.io - is there a way that this community recommends proceeding with nightwatchjs?
hey @Victoria0214, IIUC, you should be able to use .containsText with a custom message via the optional msg parameter - https://nightwatchjs.org/api/#assert-containsText.
Hello all, I'm new to nightwatch... i use the alpha.4. I try to launch chrome with an unpacked extension or firefox with a packed extension in zip format. For chrome, I try with the "arg: ['--load-extension=/path/to/my/extension/code']" in the conf file but it does not work (similar config works with puppeter or php-webdriver). For Firefox, I'm trying to use firefox.installAddon... but I must do something wrong and I could not find an example... Could someone points me to the right direction ? Thx
Hello @gravityvi, I start with the auto-generated one so I put only the chrome part :
chrome: {
desiredCapabilities : {
browserName : 'chrome',
'goog:chromeOptions' : {
// More info on Chromedriver: https://sites.google.com/a/chromium.org/chromedriver/
//
// This tells Chromedriver to run using the legacy JSONWire protocol (not required in Chrome 78)
w3c: true,
args: [
'--load-extension=/home/john/Dev/Liberty/libertyGC'
]
}
},
webdriver: {
start_process: true,
server_path: '',
cli_args: [
// --verbose
]
}
},
I try w3c: false/true... does not matter
So when I read the news on the blog saying that new alpha implements something like browser.installAddon, I give it a try full of hope ;)
Hey, I'm trying to figure out the expected behaviour for "abortOnAssertionFailure" when using Async/Await.
when i'm using ES6 Async Await and calling "waitForElementVisible" if the element is not found, the test continue to execute, the function is not throwing an Error to stop the test.
when i'm using ES6 Async Await and calling "waitForElementVisible" if the element is not found, the test continue to execute, the function is not throwing an Error to stop the test.
if i'm using the regular syntax
browser.waitForElementVisible(selector, timeout); it work as expected, fails the test and skip the other ones
when using
await browser.waitForElementVisible(selector, timeout); the test continue and ignore the flag of 'abortOnAssertionFailure'
the test will still fail, but will continue to run until it's done. and it's not my intention
@johndegey try using capabilities like this:
const chrome = require('selenium-webdriver/chrome');
const capabilities = new chrome.Options();
capabilities.addExtention(<file-path>)
module.exports = {
test_settings: {
chrome: {
capabilities,
webdriver: {
start_process: true,
server_path: require('chromedriver').path,
cli_args: [
// --verbose
]
}
}
}
};
@eliorcohen seems to me like a bug could raise an issue regarding this on github?
@gravityvi just found some old bug that may be related (i guess that waitForVisible using the assert functionality) - nightwatchjs/nightwatch#2535 , it's a year old and still open
3 replies
i't trying a simple solution wo wrap the command and handle the async result
function waitForElementVisible(selector, timeout = 1500) {
return new Promise((resolve, reject) => {
console.log(`elementIsVisible:${selector}`);
b.waitForElementVisible(selector, timeout, (data) => {
if (data.status === -1) reject(data.result.error);
resolve();
})
});
}
this is something that seems to be built-in the engine, it's check for ES6 and Async in the _watFor and such and can throw this error instead of returning a valid result for status inspection
@gravityvi Hi, nothing seems to work, (type => capabilities.addExtensions), I delete the config file and regenerate it again and try simple change : can't even start in headless mode or maximized... no options are reflected in the command line use . If I use the logged command line from chromedriver.log, add --load-extension=/my/path/to/ext/folder --start-maximized, everything works fine. Also, the autogenerated file use the entry "goo:chromeOptions" in chrome.desiredCapabilities and use "chromeOptions" in 'selenium.chrome'.desiredCapabilities, this is confusing...
I try to set chrome binary but it is always using the same version, the 'chromeOptions - binary' does not work
could be great to have a complete conf file with all options as an example
I'm pretty confident Nightwatch is able to start a chrome browser by now...
over all these years, you'd imagine that the team would have figured that one out by now
here's how I do it:
npm install chromedriver
nightwatch --env chromeit never fails.
for the default and that the config chrome.webdriver did not override this....
2 replies
It only 2 days I play with Nightwatch (v2 alpha), I did not notice the difference about the server_path :)
making a default only with chrome settings works