This problem is particularly acute on Firefox.

Hi, 

I am experiencing a similar issue in Internet Explorer with IE8. (latest 2.0b2 webdriver)
http://www.patrickacarrell.com/
web page does not completely load in IE most of the time. Therefore my code does not
continue and the page is infinitely displayed. No exceptions are displayed.

What would the best option be to handle this situation.

Thanks,

Risko

Hello,

i've got also the same issue and one wek now i am trying to find a solution for that
with no any luck.

Firefox used to have a config option (http:connection-timeout) for this issue but not
anymore. They say that this problem is at OS level and not of the browser's.

So i started  looking for a fix at WebDriver, after to Firefox and now at OS, but till
now nada.

Is there anyone out there with the same issue who managed to handle it succesfully?
This problem really kills all my tests! 

Any tip will be much appreciate. 

Thank you,
Dimitris

Hopefully this improvement will be added soon! ...because if page never finished loading
we will sit there twiddling our thumbs indefinitely!

Here is a work-around which _may not_ work for everyone.
Idea is simple:

1) Right before 'get' operation, start new thread which will be our timer
2) Perform 'get' operation
3) If thread timer expires, it will hit escape which will halt loading of that page
(you can even be more destructive if you like, like closing the browser window)
4) After either ESC caused page to stop loading or page loaded successfully 'get' will
complete operation
5) Interrupt the thread in-case it is still ticking
6) Check to see if thread had returned successful or not (use a static object)

Attached sample code: webdriver-timeout.zip

Oh thanks a lot ) I hope this hit will work in chrome )

Many Thanks!

Just one thing with the robot interface. Is it true that works only when the FF window
is on focus? I suppose this will be more tricky when we got several test running in
paralel..

I hope this feature will be add it in the next releases..


Thank you again! 

Dimitris, 

I had the same thing (multi-threaded tests w/ 3 instances of each browser)... but I
just tweaked mike's example to also pass in the instance of WebDriver to the thread
and in the TimeoutThread I stop the window with javascript:
((JavascriptExecutor) wd).executeScript("window.stop();");
I also am taking a screenshot for debugging purposes.

Hi Luke,

thanks for the tip! I also tried to stop page loading by executing the window.stop()
command with a setTimeout()  before loading the page, but it didnt work. Probably this
was because of the order that script are running, as in mozila website say:

Because of the order in which scripts are loaded, the stop() method cannot stop the
document in which it is contained from loading, but it will stop the loading of large
images, new windows, and other objects whose loading is deferred.

Anyway i am gonna try your tip, since to kill the browser is not the best option...

Cheers

Sorry... my suggestion doesn't work for all (many) cases it seems for me.

So, instead I patched Firefox driver to include a timeout for page loading (sorry,
my main problem is w/ firefox, so i didn't try to patch any others). Here's my first
stab at it attached. Default timeout is 15 min (but you can override by using a FirefoxProfile
and calling setPageLoadTimeout)

Issue 221 has been merged into this issue.

For python:

urllib2 accepts a timeout variable, but the Se Python bindings don't currently allow
access to it.

My solution is to import socket and then put this in the setUp method:
    socket.setdefaulttimeout(30)

Issue 2618 has been merged into this issue.

For the Firefox users out there, instead of hoping for the patch i provided, Simon introduced
a 'fast loading' switch that can be used. This simply changes the behavior of firefox
to not wait for everything to load, so you have to be careful that most things will
likely not be on the page and you should use and Explicit or Implicit wait when finding
(anything).

http://code.google.com/p/selenium/wiki/FirefoxDriver#-Beta-_load_fast_preference

Is there an ETA as to when the beta 'fast loading' switch will be available specifically
for .Net?  I'm assuming that since you posted today it is not available yet and I didn't
see anything in the change log at http://selenium.googlecode.com/svn/trunk/dotnet/CHANGELOG

It is available (and has been for a while). You just have to do the C# equivalent for
a FirefoxProfile. I only just now posted because someone asked in the IRC channel today.
I want to say this has been available since 2.15 or maybe earlier. 

Thanks Luke!  I'll give it a whirl in the current version I have (2.14.0).  If not I'll
upgrade my Selenium binaries to the latest.

Just looked up that the load strategy preference was first available in 2.9

In 2.19 the value that will need to be passed in is changing from 'fast' to 'unstable'
 (wiki page has been updated accordingly).

An end point for generic timeouts has been set up in r15880. This is where a new "page
load" timeout can be added.

This page load timeout functionality has been hooked into the firefox driver and will
be available in 2.20. Note that you'll also need to use the "unstable" page loading
strategy for your tests to run without muss or fuss.

So there is still no way to have a timeout with the "normal" page loading strategy?
 The "unstable" strategy will return from a driver.get() with some arbitrary amount
of content on the page not actually loaded, even when there is no timeout, right? 
That seems not nearly as useful...

Hello, all.
I have this problem. I go to url where content is very long time loading and webDriver
bad work (I want get screen shot). I have a questions, maybe someone knows when this
bug (bug: can't set property pageLoadTimeout in FirefoxDriver) is resolved?  Thanks.
Oh, yes, method where change page loading strategy don't work (my WebDriver is waiting
full page loading). Thanks. 

I just tried this with selenium 2.20 in remote webdriver and grid2.
The change log says: " * Introduce a timeout for page loads. This needs to be used
in
    conjunction with the unstable page load detection."
I have issue with the timeout set by
 driver.manage().timeouts().pageLoadTimeout(30000,TimeUnit.MILLISECONDS);
 when I use it together with the FF profile
 ffProfile.setPreference("webdriver.load.strategy", "unstable");
Without the unstable option it seems good. I set the timeout to 300ms and I got the
pageload timeout. I set 30000ms and the everything just worked fine. The change log
should be more explicit by telling do not use them together or something, because telling
that "in conjunction" is similar to write "use it properly" which hold no information
but people end up in confusions :)
I would be happy to see this behavior for other browsers too.
Thank you for this feature!

I have tried with 2.20, remote webdriver and grid2 as well but I cannot execute my test.
When loads the configuration in @BeforeTest function, returns an error in execution
time: java.lang.NoSuchMethodError: org.openqa.selenium.WebDriver$Timeouts.pageLoadTimeout(JLjava/util/concurrent/TimeUnit;)Lorg/openqa/selenium/WebDriver$Timeouts;

If I remove the line driver.manage().timeouts().pageLoadTimeout(30,TimeUnit.SECONDS);
it works properly. How can I set pageLoadTimeout without errors?
Regards

This seems to be a very common problem, and certainly is for us.

I am facing similar issues in Selenium Webdriver [Python]. It would be kind enough if
someone fixes this problem. Well, it is a fundamental problem with many tests. Thank
you.

Is there any plan to fix this?  I am running into this with paypal's sandbox checkout
system when entering email/password.  Everything I have found on this topic has not
seemed to workl

I am also facing same timeout issue(The HTTP request to the remote WebDriver server
for URL http://localhost:7055/hub/session/xx-xx-x-x-x/element/%xx-xx-x-x-x%7D/click
timed out after 60 seconds) with firefox randomly while clicking a fileupload button
or a link.. Its so much frustrating. Same test sometimes work fine if I revert back
to old FF or old Selenium driver. Currently I am using webdriver 2.41 and FF28.

Hello, I'm having the same problem with FF 28 and selenium 2.41. The test hangs randomly
waiting for browser response. The only way I have to recover this error is closing
the browser, but I can't do it from my test because the test is hanged.
Even in the pages that takes long time to load, the timeout property is not working,
the test will wait until the page is fully load. I think that the desired behaviour
will be throwing a timeout exception after the time indicated is consumed.

This problem means that the test will hang forever, doing the test useless.

The explanation and stacktrace is here: http://stackoverflow.com/questions/22933913/firefox-webdriver-hangs-waiting-for-page-to-load

On #62, if this issue isn't fixed, an interim workaround could be to create a monitor/watcher/listener/poller
that checks for the browser process and if it's still running after x amount of time,
kill it so test won't hang. Such a monitor process would have to be separate from the
test being run. Either it is forking a new monitor process from the test during execution,
or it's separately started/invoked outside of the test.

Every 6 months or so I look into moving from Selenium Server to Web Driver and fail
miserably because of this issue.  What is the point of having Selenium Timout and Implicit
when they don't actually do anything.  We test performance, using Wait Until Page contains
Button/Element, etc emulating impatient users who don't want to the 200 minutes for
ads.jiwire.com to complete.  I still have no idea what the wait until keywords would
actually do since you have to have page completely loaded before the open browser or
click element exits.  

I am facing this issue in an applications' login page. Nothing happens even if I wait
for 15 minutes.. no errors or information..

Versions: IE 9, Selenium 2.35

Is there any Java way of handling or forcing the browser to finish its loading and
make the tests to run?

i just added an extra wait function to make WebDriver wait page to load
protected void wait(int seconds) {
        try {
            Thread.sleep(1000 * seconds);
        } catch (InterruptedException e) {
        }
    }

I believe I'm having a similar issue with Selenium for Java v 2.42.0. The problem seems
to be happening with Firefox 28 and 29. Selenium calls (findElement, findElements,
isDisplay) will hang for no apparent reason. No timeout, no error messages thrown VERY
frustrating and a major hang up for me. Has anyone else experiences this?

I have been facing this issue quite recently. WebDriver hangs upon calling FindElement/Click().

I have been using FF23 and webDriver for Java v 2.39. Any fix?

Can someone at least shed light on reason of this failure?

FYI, if anyone is experiencing this issue from clicking a link that causes a page load
to new page, etc., and you're using WebDriver's native WebElement click() method, a
workaround may be to click the element from javascript. I ran into such an issue and
switching to javascript didn't cause the page load hang.

While I love Selenium and PhantomJS, this is my number one obstacle to using it and
a source of great frustration. 

I am also facing the same problem (using Chrome and Python) and there is no way to stop
the page reload. There is a similar problem (Issue 6867) and its marked as fixed but
using pageLoadTimeouts() doesn't help as there are lots of ads in the page and they
keep on loading the page.

On comment #73, sounds like a possible workaround for you could be to route your browser's
traffic through a (BrowserMob) proxy. And configure proxy to filter out the ads, so
that the page will then load (w/o the ads or broken images, etc. where the ads go).

I also facing the problem that Firefox hangs up on a certain page. I am using Selenium
python webdriver wrapper (2.42.0 - also tried 2.43) on Ubuntu 14.04 on Virtual Box
(3.6.4) running on Windows 7 (64bit). I discovered that it hangs up when the network
connection on Virtual Box is set NAT, it works well when it is set to Bridge. I think
the reason might be that the website by default loads itself as https. Although, I
am not certain of it. I am only certain that it worked for me on Bridged connection
and does not work at all through NAT.

To deal with the ads I installed http://www.privoxy.org/ (much easier and faster to
configure than Squid). 

I'm having this issue as well where a test will hang forever using Firefox 24.8.1 ESR
and Selenium 2.43.1. It looks like there needs to be a timeout set when trying to execute
an HTTP request. Apache HTTP library supports a timeout but it doesn't look like it
is being set. Here's the stack trace:

  [junit] "Forwarding findChildElement on session 6e7c8385-71c9-4eee-945a-44113d3dcd17
to remote" prio=10 tid=0xb660b800 nid=0xf69 runnable [0xb67e6000]
    [junit]    java.lang.Thread.State: RUNNABLE
    [junit]     at java.net.SocketInputStream.socketRead0(Native Method)
    [junit]     at java.net.SocketInputStream.read(SocketInputStream.java:152)
    [junit]     at java.net.SocketInputStream.read(SocketInputStream.java:122)
    [junit]     at org.apache.http.impl.io.SessionInputBufferImpl.streamRead(SessionInputBufferImpl.java:136)
    [junit]     at org.apache.http.impl.io.SessionInputBufferImpl.fillBuffer(SessionInputBufferImpl.java:152)
    [junit]     at org.apache.http.impl.io.SessionInputBufferImpl.readLine(SessionInputBufferImpl.java:270)
    [junit]     at org.apache.http.impl.conn.DefaultHttpResponseParser.parseHead(DefaultHttpResponseParser.java:140)
    [junit]     at org.apache.http.impl.conn.DefaultHttpResponseParser.parseHead(DefaultHttpResponseParser.java:57)
    [junit]     at org.apache.http.impl.io.AbstractMessageParser.parse(AbstractMessageParser.java:260)
    [junit]     at org.apache.http.impl.DefaultBHttpClientConnection.receiveResponseHeader(DefaultBHttpClientConnection.java:161)
    [junit]     at sun.reflect.GeneratedMethodAccessor7.invoke(Unknown Source)
    [junit]     at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
    [junit]     at java.lang.reflect.Method.invoke(Method.java:606)
    [junit]     at org.apache.http.impl.conn.CPoolProxy.invoke(CPoolProxy.java:138)
    [junit]     at com.sun.proxy.$Proxy5.receiveResponseHeader(Unknown Source)
    [junit]     at org.apache.http.protocol.HttpRequestExecutor.doReceiveResponse(HttpRequestExecutor.java:271)
    [junit]     at org.apache.http.protocol.HttpRequestExecutor.execute(HttpRequestExecutor.java:123)
    [junit]     at org.apache.http.impl.execchain.MainClientExec.execute(MainClientExec.java:254)
    [junit]     at org.apache.http.impl.execchain.ProtocolExec.execute(ProtocolExec.java:195)
    [junit]     at org.apache.http.impl.execchain.RetryExec.execute(RetryExec.java:85)
    [junit]     at org.apache.http.impl.execchain.RedirectExec.execute(RedirectExec.java:108)
    [junit]     at org.apache.http.impl.client.InternalHttpClient.doExecute(InternalHttpClient.java:186)
    [junit]     at org.apache.http.impl.client.CloseableHttpClient.execute(CloseableHttpClient.java:72)
    [junit]     at org.apache.http.impl.client.CloseableHttpClient.execute(CloseableHttpClient.java:57)
    [junit]     at org.openqa.selenium.remote.HttpCommandExecutor.fallBackExecute(HttpCommandExecutor.java:204)
    [junit]     at org.openqa.selenium.remote.HttpCommandExecutor.execute(HttpCommandExecutor.java:173)
    [junit]     at org.openqa.selenium.firefox.internal.NewProfileExtensionConnection.execute(NewProfileExtensionConnection.java:165)
    [junit]     at org.openqa.selenium.firefox.FirefoxDriver$LazyCommandExecutor.execute(FirefoxDriver.java:362)
    [junit]     at org.openqa.selenium.remote.RemoteWebDriver.execute(RemoteWebDriver.java:572)
    [junit]     at org.openqa.selenium.remote.RemoteWebElement.execute(RemoteWebElement.java:268)
    [junit]     at org.openqa.selenium.remote.RemoteWebElement.findElement(RemoteWebElement.java:171)
    [junit]     at org.openqa.selenium.remote.RemoteWebElement.findElementByCssSelector(RemoteWebElement.java:236)
    [junit]     at org.openqa.selenium.By$ByCssSelector.findElement(By.java:426)
    [junit]     at org.openqa.selenium.remote.RemoteWebElement.findElement(RemoteWebElement.java:167)
    [junit]     at XXXXX.ControllerHelper.getNameOfActiveContoller(ControllerHelper.java:301)
    ...

Comment #74 would have helped but the problem is I *need to* work with ads. I am working
on a project which requires the ads to be in the web page. And when the ads load selenium
web drivers becomes no help as the page hangs up. All I need is the page source after
some time. Whenever I try to get the page source I get the TimeoutException. Isn't
there any way to get the (partial) page source and close the irresponsive tab after
I know that the page is taking forever to load?

#61 could somebody tell me which old FF version and which old selenium driver works
fine with page load timeouts? i might need an old FF for my solution so that would
not be a problem.

Mostly implemented in Chrome

Issue 6015 has been merged into this issue.

Our app is built in .NET. I'm doing Selenium/Java using Eclipse. 
I use thread and timeout together to go around this issue. And it works for Chrome.
Thread.sleep(5000);
TimeUnit.SECONDS.sleep(20);
I couldn't find any other way. 
Specially we have popup window to be loaded and clicked. I used this timeout settings
when I have to click the popup window. Not the best solution, but works ok. 

Re: #71 , executing element.click() in Javascript did not help in un-deadlocking a Safari
6 web driver on OS X 10.0.6 (I guess the comment's author addressed a different issue
where the click call itself hung).  Our web UI's element click handler executes window.open(URL),
and attempts to test the pop-up window with any WebDriver call such as getTitle() causes
that call to deadlock until reaching the socket timeout.

As a side note, I was surprised to observe our tests pass with the javascript element.click()
even on Internet Explorer which used to block javascript-initiated pop-ups.  I guess
the test passed because it unexpectedly relied on a cookie from a previous manual session.

(Earlier I worked around the IE blocker by calling WebElement#click() against an element
with a click handler that executed window.open()).

Curious on #83, what is visually observed when the test blocks? Does the popup window
come up? Does the content of the popup load in the popup window? What happens if you
manually close the popup, will that trigger a faster (failure) response in the test?

Hi All,

Facing timeout issue while calling firefox via webdriver but when I opened it manually
and entered same URL , got opened successfully within a minutes.
Could you please suggest any solution