While running automation tests I sometimes get a timeout error for Selenium Webdriver (I think this is where the issue is at least). Me and my team have all recently migrated to Macbooks (from a combination of Windows and Ubunutu machines) and are all getting this behaviour.
While running a suite of tests I will (seemingly at random) get the following error output in the console:
Errno::ETIMEDOUT: Failed to open TCP connection to 127.0.0.1:9515 (Operation timed out - connect(2) for "127.0.0.1" port 9515)
This doesn't happen consistently, sometimes I'll run a pack and not have any such errors, sometimes I'll have multiple occurrences.
Here is the code which registers the driver (in case anything here points to what the issue could be):
Capybara.register_driver :selenium do |app|
opts = Selenium::WebDriver::Chrome::Options.new
opts.add_argument '--start-maximized'
opts.add_argument 'disable-infobars'
opts.add_argument '--disable-notifications'
opts.add_preference(:safebrowsing,
enabled: true)
opts.add_preference(:browser, set_download_behavior: { behavior: 'allow' })
Capybara::Selenium::Driver.new(app, browser: :chrome, options: opts)
end
The gems I am using are Capybara (3.11.0), Cucumber (3.1.0) and Selenium-webdriver (3.141.0). I have ChromeDriver(73.0.3683.68) installed via HomeBrew
Has anyone encountered this issue and worked out what the cause is?
Port 9515 is the default port chromedriver runs on. If you happen to be using Chrome/chromedriver v74 try rolling back to 73 or forward to 75 - 74 has been reported to have issues where it will randomly hang.
Another potential solution is to upgrade to Capybara >= 3.16.0 which defaults to using a persistent connection to chromedriver. This would mean less opening/closing of connections and less chance for chromedriver to hang during connection establishment.
Related
# encoding: UTF-8
# language: ru
require 'webdrivers'
require 'selenium-cucumber'
options = Selenium::WebDriver::Options.chrome
driver = Selenium::WebDriver.for :chrome, options: options
#driver.get "https://www.google.com"
Given 'Переходим на сайт' do
driver.get "https://www.google.com"
end
sleep(2)
driver.quit
File .feature
# encoding: UTF-8
# language: ru
Функция: Тест
Сценарий: Тест тест
Дано Переходим на сайт
The code is as simple as possible. And it works if you remove the Given function and start opening the site not through cucumber. Everything works and opens. But if you open it through Given, an incomprehensible error with ports appears.
Failed to open TCP connection to 127.0.0.1:9515 (Connection refused - connect(2) for "127.0.0.1" port 9515) (Errno::ECONNREFUSED)
./features/step_definitions/test_steps.rb:23:in `nil'
Linux Mint
This is likely due to lazy initialization in Selenium. When you call driver.get "https://www.google.com" directly it's initializing the selenium driver and browser connection, so when driver.quit is called the connection is up and can then be cleanly shut down. However, when you call
Given 'Переходим на сайт' do
driver.get "https://www.google.com"
end
you are defining a block that can be called later, but it doesn't actually execute the driver.get ... yet, so the driver to browser connection isn't fully setup yet which leads to quit being called on a non-initialized connection and therefore raising an error.
I am using Capybara Selenium to run headless Chrome, which works great, except I cannot figure out how to use remote debugging. When I add --remote-debugging-port=4444 or --remote-debugging-port=9222 or --remote-debugging-port=9521, Selenium no longer connects to the browser to run the test.
How do I get remote debugging to work? Here is my code for reference:
Capybara.register_driver :selenium do |app|
# from https://github.com/SeleniumHQ/selenium/issues/3738
capabilities = Selenium::WebDriver::Remote::Capabilities.chrome(loggingPrefs: {browser: 'ALL'})
options = Selenium::WebDriver::Chrome::Options.new
options.add_argument '--disable-infobars' # hide info bar about chrome automating test
# if we don't use this flag, every selenium test will die with the error:
# "unknown error: Chrome failed to start: exited abnormally"
options.add_argument '--no-sandbox'
# BREAKS THINGS if uncommented
# options.add_argument '--remote-debugging-port=4444'
options.add_argument '--headless'
options.add_argument '--window-size=1600,2400'
options.add_preference('profile.default_content_settings.popups', 0)
options.add_preference('download.default_directory', DownloadHelpers::PATH.to_s)
Capybara::Selenium::Driver.new(
app,
clear_local_storage: true,
clear_session_storage: true,
browser: :chrome,
options: options,
desired_capabilities: capabilities,
)
end
Since chrome 67 and chromedriver 2.39, chromedriver now correctly uses the port you specify with --remote-debugging-port. This removes quite a bit of complexity from my answer above. The steps I now take, which work for my use case of needing to configure download settings using chrome_remote, are as follows:
It makes uses of a nodejs library, crmux - which allows multiple clients to connect to the remote debug port of chrome at the same time.
Get nodejs installed first: Nodejs v9.7.0 works fine
Install crmux by running npm install crmux -g
Before you start chromedriver (Capybara::Selenium::Driver.new), you need to spawn a separate thread that will fire up crmux, which will let both you and chromedriver communicate with chrome itself via the port you specified in Capybara (4444):
crmux --port=4444 --listen=4444
You may want to add a sleep 3 after the spawn command in the main script/thread to give time for crmux to start before you continue with your test startup.
You can then use chrome_remote (for example) to access chrome using port 4444, while capybara is doing its thing.
Updating my ChromeDriver fixed it for me. I didn't have to do anything else. Before it would hang when attempting to start the test.
Specifically I was on ChromeDriver 2.36 and I upgraded to ChromeDriver 2.40. I don't think the Chrome version was a problem, since I was on Chrome 67 both before and after.
Here's how I'm registering the driver:
Capybara.register_driver :headless_chrome do |app|
capabilities = Selenium::WebDriver::Remote::Capabilities.chrome(
chromeOptions: { args: %w[headless window-size=1280,960 remote-debugging-port=9222] }
)
Capybara::Selenium::Driver.new(app, browser: :chrome, desired_capabilities: capabilities)
end
After that I ran my test with a debugger (binding.pry) placed where I wanted to inspect. Then when I hit the debugger I navigated to http://localhost:9222/ in a normal instance of Chrome and was able to follow a link to view what was happening in the headless Chrome instance, including the browser console output I needed.
Update: if using versions since Chrome 67/chromedriver 2.39, my alternative answer above provides a simpler solution
The core issue here is Chromedriver also uses the remote debugging port connection to communicate with Chrome. This uses the websocket protocol, which only supports a single-client connected at a time. Normally, when chromedriver starts the chromedriver process, it will choose a random free TCP port number and use this to access the remote debug port. If you specify --remote-debuggging-port=9222 however, Chrome will be opened with the debug port you have requested, but chromedriver will silently continue to try and open a connection using this random port number.
The solution I ended up with was heavily inspired by comment 20 in this chromedriver issue. It required quite a bit of code to get it working, but works solidly. It makes uses of a nodejs library, crmux - which allows multiple clients to connect to the remote debug port of chrome at the same time.
Get nodejs installed first: Nodejs v9.7.0 works fine
Install crmux by running npm install crmux -g
Before you start chromedriver (Capybara::Selenium::Driver.new), you need to spawn a separate thread that will do a few things: look for the remote debug port chromedriver is trying to use to connect to chrome, and then use this to fire up crmux. Once this has happened Capybara etc will work as normal.
My separate thread runs a ruby script that first executes a netstat command repeatedly until it finds the relevant entry for chromedriver (TCP status is SYN_SENT). This separate thread must continue to run in the background while chrome is up and running.
The code for this is:
$chrdrv_wait_timeout = 60
$chrdrv_exe = "chromedriver.exe"
def get_netstat_output
stdout = `netstat -a -b -n`
stat_lines = stdout.split("\n")
stat_lines
end
def try_get_requested_port
socket_state = "SYN_SENT" # i.e. sent with no reply
statout = get_netstat_output
n = statout.length
i = 0
loop do
i += 1
# find lines relating to chromedriver
exe_match = /^ +\[#{$chrdrv_exe}\]$/.match statout[i]
if exe_match != nil
# check preceeding lines which should contain port info
port_match = /TCP.*:([0-9]+)\W+#{socket_state}/.match statout[i-1]
if port_match != nil
return port_match[1].to_i
end
end
break unless i < n
end
return nil
end
def get_tcp_port_requested_by_chromedriver
i = 1
loop do
puts "Waiting for #{$chrdrv_exe}: #{i}"
port = try_get_requested_port
if port != nil
return port
end
break unless i < $chrdrv_wait_timeout
sleep 1
i += 1
end
raise Exception, "Failed to get TCP port requested by #{$chrdrv_exe} (gave up after #{$chrdrv_wait_timeout} seconds)"
end
(I'm working in Windows: for Mac/Linux the netstat syntax/output is probably different so the code will need adjustment; the key thing is you need it to output the executable owner of each connection entry - and parse the bit relating to chromedriver to get the port in question).
Once this has found the random port (I'll use 12225 as an example), the background ruby script can then execute a crmux process, which will reunite chromedriver with chrome itself via the port you specified in Capybara (4444):
crmux --port=4444 --listen=12225
Finally, this separate script saves the discovered listen port to a text file. This allows the main script/thread that is running capybara to know the port number it can use to get access to chrome (via crmux's multiplexed connection) by reading in the port from that file. So you can then use chrome_remote to access chrome using port 12225, for example, while capybara is doing its thing.
I am using firefox v 48.0.2 and am trying to get my selenium (selenium-server v2.53) remotedriver automated tests to run on firefox, I have geckodriver 0.9.0 installed and when i go through the documentation on the github readme and run this command: (on mac osx 10.11.3)
geckodriver -b /Applications/FirefoxNightly.app/Contents/MacOS/firefox-bin
i get this error message:
thread '< main >' panicked at 'called Result::unwrap() on an Err value: Io(Error { repr: Os { code: 48, message: "Address already in use" } })', ../src/libcore/result.rs:746
note: Run with RUST_BACKTRACE=1 for a backtrace.
ive tried ignoring this step but when i run my tests firefox does not launch, i have ensured that my webdriver capabilities include marrionette: true
WebDriver:
browser: 'firefox'
clear_cookies: false
restart: false
window_size: 414x736
marionette: true
other than that i can not find any documentation to lead me in the correct direction, did i perhaps over look something? any help is greatly appreciated!
I am also using codeception to handle my tests (php)
EDIT
I was able to get this command to work after killing the process that was listening on port 4444:
geckodriver -b /Applications/FirefoxNightly.app/Contents/MacOS/firefox-bin
But even with that running Firefox is still not launching
I've searched in the archive but could not find an answer to my dilemma. I'm coding in Ruby and using watir webdriver framework on my local Mac Yosemite and want to connect to postgres database on a linux box.
I have the required ruby gems installed on my local Mac
* LOCAL GEMS *
dbd-pg (0.3.9)
pg (0.18.4)
dbi (0.4.5, 0.4.4)
I am using the following code.
require 'rubygems'
require 'pg'
require 'dbd/pg'
require 'dbi'
conn = PGconn.connect("10.0.xx.xx","5432",'','',"mydbname","dbuser", "")
res = conn.exec('select * from priorities_map;')
puts res.getvalue(0,0)
conn.close if conn
On running this
I a getting these errors
.initialize': Could not connect to server: Connection refused (PG::ConnectionBad)
Is the server running on host "10.0.xx.xx" and accepting
TCP/IP connections on port 5432?
If I use the code
dbh = DBI.connect("dbi:pg:mydbname:ipaddress", "user", "")
row = dbh.exec('select * from etr_priorities_map;')
puts row.getvalue(0,0)
dbh.disconnect if dbh
I get the error
block in load_driver': Unable to load driver 'pg' (underlying error: wrong constant name pg) (DBI::InterfaceError) from System/Library/Frameworks/Ruby.framework/Versions/2.0/usr/lib/ruby/2.0.0/monitor.rb:211:in `mon_synchronize'
I am new to Ruby. How can I resolve these issues?
The first error, as #Doon said in the comments, comes from the TCP connection and usually means your database is not listening on the network. Most PostgreSQL packages come with a default configuration that only allows local connections, but you can enable connections over the network in the server configuration via the listen_addresses setting. I installed PostgreSQL through Homebrew on my Mac, and my config is at /usr/local/var/postgres/postgresql.conf, but if you installed it some other way the path may be different.
The second error is happening because the "driver" part of the connection string is case-sensitive, and the DBD driver for Postgres is named Pg, not pg. Try this:
dbh = DBI.connect("dbi:Pg:mydbname:ipaddress", "user", "")
Also, unless you have your heart set on using Ruby/DBI, you might want to consider using a more-recently maintained library. Ruby-DBI is very well-written and tested, but it hasn't seen a release since 2010, and Ruby itself has changed pretty significantly in the interim.
If you do want to consider alternatives, I use Sequel for mostly everything, and I highly recommend it, especially for Postgres development, but both DataMapper and ActiveRecord have a large userbase as well.
I want to use SSL with MongoDB. It's not enabled by default so one has to compile from source with the necessary options. I followed the official documentation and got the v2.6.4 binary built and running nicely on a freshly deployed server running Ubuntu 14.04. All good so far.
Next I set up mongod as described in the official docs. I did follow their example of using a self-certified key for testing purposes. And the relevant part of the config looks like:
...
net:
bindIp: 127.0.0.1
port: 27017
ssl:
mode: requireSSL
PEMKeyFile: /opt/mongo/security/mongodb.pem
...
If I then run the client and specify to use SSL I connect fine. ($ mongo --ssl). FWIW if I try without the --ssl argument then it doesn't connect.
Ok, time to link up via Ruby. I'm on the same server and I try the following ruby script:
require 'rubygems'
require 'mongo'
client = Mongo::MongoClient.new('localhost', 27017, {:ssl => true})
Nope. It's not having it:
/home/test/.rvm/gems/ruby-1.9.3-p547/gems/mongo-1.11.1/lib/mongo/mongo_client.rb:422:in `connect': Failed to connect to a master node at localhost:27017 (Mongo::ConnectionFailure)
from /home/test/.rvm/gems/ruby-1.9.3-p547/gems/mongo-1.11.1/lib/mongo/mongo_client.rb:661:in `setup'
from /home/test/.rvm/gems/ruby-1.9.3-p547/gems/mongo-1.11.1/lib/mongo/mongo_client.rb:177:in `initialize'
from test_mongo_ssl.rb:8:in `new'
from test_mongo_ssl.rb:8:in `<main>'
So best to make sure that there's nothing wrong with the default connection without SSL. I disable SSL on mongod and restart. Then try the ruby script again, this time without the ssl option:
...
client = Mongo::MongoClient.new('localhost', 27017)
And it's fine. Therefore I feel I've narrowed it down to the ruby driver & ssl, but beyond that there's little else to go on.
EDIT I tried their Python driver on the same server and used their example program:
from pymongo import MongoClient
c = MongoClient(host="localhost", port=27017, ssl=True)
And that did connect OK. So at least I can feel fairly confident that the mongod is configured properly and the issue lies somewhere within the Mongo Ruby driver. Quite possibly a bug in their current driver (v1.11.1).
UPDATE I've also had success connecting via ssl using the node.js driver:
var mongo = require('mongodb');
var database = new mongo.Db("my_database", new mongo.Server("127.0.0.1", 27017, {ssl:true} ), {w:0});
database.open(function(err, db) {
if(err) throw err;
db.authenticate('user', 'password', function(err, result) {
var collection = db.collection('foo');
collection.findOne(function(err, item) {
if(err) throw err;
console.log(item);
db.close();
});
});
});
There it seems to be increasingly likely that there's either a bug in the ruby driver, or the documentation is incomplete and not explaining accurately how to use SSL connections. Therefore I've opened a new issue on MongoDB's issue tracker to hopefully get to the bottom of this.
Rather embarrassingly, the solution to this issue was my /etc/hosts file had a typo for the localhost entry:
127.0.0.1 localhost.localdomain locahost
As you can see, it's missing the second letter L in "localhost". (I suspect it went missing during an accidental vim gesture.) Therefore to resolve I just had to reinstate the missing "l":
127.0.0.1 localhost.localdomain localhost
It's still a mystery as to why the Python sample worked correctly. And it's because of that I didn't twig earlier that it was a problem with the hosts file.