Closed lukeis closed 8 years ago
In order to be able to reproduce and fix the issue, need to know the version of Selenium
you are using, the code you are executing when the error happens, and either an HTML
page (along with the appropriate CSS and JavaScript files) or a public URL to a site
that reproduces the issue. This issue is not encountered by every piece of code that
uses the IE driver, so narrowing down the cause would be our first step.
Reported by james.h.evans.jr
on 2011-11-28 11:23:49
NeedsClarification
Add more info so that the issue can be located. attached pls find the error output
Reported by ruby2001
on 2011-11-29 02:31:59
Reported by barancev
on 2011-11-29 18:35:32
I find the same problem. After some investigation i find that this error happens when
you create and destroy InternetExplorerDriver in different threads. I created simple
maven project, that you can use to reproduce this issue.
Also it doesn't metter you run this on virtual machine or on standalone Windows machine.
If you need more information about this problem don't hesitate to aks.
Reported by alexandr.kulik1988
on 2011-11-30 12:43:20
Also i found if you run maven with -DformMode=never everything works fine. So my assumption
about different threads was wrong.
Reported by alexandr.kulik1988
on 2011-11-30 13:56:47
I encountered same issue. Any resolution for this?
Reported by thavva1023
on 2012-01-24 10:10:31
I have also encountered similar issue.
When I'm running test remotely in virtual machine on Windows XP, the 'Java.exe has
encountered a problem' dialog pop up. Please see the attached detail and some information
below. The test can execute successfully in Virtual Machine with Firefox and Chrome.
What steps will reproduce the problem?
After successfully login into our application it comes to the main page which cause
the “java.exe has encountered a problem…”. This page requires the Chrome Frame installed
to work correctly.
Selenium version: 2.15.0
OS: Windows XP
Browser: InternetExplorer
Broswer version: 8.0.6001
Google Chrome Frame - 17.0.963.46
Reported by cindyhchiang2
on 2012-02-17 19:52:14
We still need an executable test case to fix this.
Reported by dawagner
on 2012-02-28 01:03:06
Here is what I found:
1) the jvm appears to crash when shutting down a session: driver.quit(); after some
level of interaction with web pages and closing all open pages.
2) The jvm does not always crash running a particuar test. The best prevention appears
to be removing the temp iedriver.dll created in C:\Documents and Settings\{username}\Local
Settings\Temp\webdriver*libs between test runs.
Selenium version 2.18
Java 6.22
OS XP sp 3
Browser IE 6
Reported by tdalan6009@yahoo.com
on 2012-04-02 21:42:55
I'm experiencing the same problem but couldn't isolate the part that is causing it.
Application has some popups and when the test clicks close on one of those, the error
usually occurs there. I've created a similar small application just copying the popup
opening/closing code but the problem is not reproducible on it. Will still be trying,
just please don't close the issue until it's resolved...
Selenium version 2.20
Java 6.31
OS XP Sp3 (VMware)
Browser IE 6
Reported by aleksander.wirth
on 2012-04-03 06:37:56
I ran into a similar issue. I have not been able to isolate it. However, I did find
a work-around. I had a fairly long-running test sequence. I found that if closed
and reopened my driver at convenient points in the sequence, I could avoid the problem.
Selenium version 2.20
Java 1.6.0_19
OS XP Sp3 (VMware)
Browser IE 8
In my case, the test code is being executed via a RemoteDriver.
Reported by houseofyin
on 2012-04-04 16:33:39
I have had similar issues,
* Remote WebDriver via java
* Selenium server standalone jar (2.21)
* Java 1.6 update 30 32bit
I do a series of 'driver.get()' URLs
At the end of testing I have been using 'driver.quit()' after all 'driver.gets()' have
been run.
Reported by brad.latus
on 2012-05-07 02:38:35
Extra note:
This sort of error generally occurs after 1+ hours of running an IE test every 5 minutes.
Reported by brad.latus
on 2012-05-07 02:51:07
Finally I have managed to get a simple script that in my tests is able to quickly kill
a remote webdriver instance. The ruby script is attached.
It looks like the server dies when it tries to collect the hung browsers that the script
intentionally leaves there (no quit calls from the script).
The webdriver server instance is started with the following config:
java -jar selenium-server-standalone-2.21.0.jar -timeout 20 -port 4444 -maxSession
1 -browser "browserName=internet explorer,version=8,maxInstances=1" -browser browserName=chrome,version=18,maxInstances=10
-Dwebdriver.chrome.driver="C:\Documents and Settings\tuenti\My Documents\chromedriver_win_18.0.1022.0.exe"
Internet Explorer version is: 8.0.6001.18702
OS: Windows XP 5.1 (service pack 3)
Reported by davidst
on 2012-05-14 15:10:14
Also, I'm running the script locally on my MacBook and the server is running inside
a VirtualBox Windows instance
Reported by davidst
on 2012-05-14 15:11:30
Just a heads-up on this: I'm seeing that the embedded http server (mongoose) that the
C++ is using is stopped (and destroyed) when the first browser is reclaimed by the
SessionCleaner thread.
Also, when the second browser is then reclaimed by SessionCleaner, as the embedded
http server is down the request fails, aaaaand the SessionCleaner thread dies with
the runtimeexception that is thrown, so no more sessions are reclaimed... ever. This
second thing is more suited to a different bug, but I just wanted to comment it as
I have come across it while debugging the original jvm crash problem.
Reported by davidst
on 2012-05-17 16:14:09
Great debugging work. I just started running into this and was going to try to figure
out what is going on. I need to get a debug build together first.
Reported by nirvdrum
on 2012-05-17 17:47:20
Reported by nirvdrum
on 2012-05-17 17:47:39
Accepted
Nevermind. I'm now testing a possible fix, but need to give it some more time and love
:)
Reported by davidst
on 2012-05-17 18:02:20
Kevin, after talking to Jim Evans last night he confirmed that some changes have been
pushed to trunk this week in the direction of using the IE driver in a way similar
to the Chrome one: and external .exe referenced with a config param that enables the
execution of an individual server for every session, so I will test how this works
with my current test scenario.
So far I was using 2.21 version.
Reported by davidst
on 2012-05-18 07:57:27
David,
I'd be interested in hearing if things clear up for you. I've been building from trunk
and am still seeing the problem, unfortunately.
Reported by nirvdrum
on 2012-05-18 12:36:23
So far I can confirm you that the server crash I was able to reproduce with the ruby
script I provided above and the way I was starting the server is not happening with
current trunk for me.
Nevertheless, I want to test it also with our current webdriver suite that is working
like a charm with Chrome.
Reported by davidst
on 2012-05-18 15:47:56
Should be fixed in 2.22, as long as you're using the standalone IEDriverServer.exe
Reported by barancev
on 2012-05-29 20:36:15
Fixed
Great - but how do i use the standalone IEDriverServer.exe when running as a grid node?
Reported by christoph.neuroth
on 2012-06-27 08:33:13
You have to pass an additional parameter when launching the .jar:
-Dwebdriver.ie.driver=c:\PATH\TO\IEDriverServer.exe
Reported by davidst
on 2012-06-27 09:03:26
Thanks. Does not help though. I can see from the log that the standalone driver server
is used, but as before, after about an hour of happily running tests, it fails with
a VC++ Runtime Error :(
Reported by christoph.neuroth
on 2012-06-27 15:00:41
I'm running into the exact same problem as christop...@gmail.com here is our log file.
I've been trying to resolve this for months :(
Reported by elderjameskieley
on 2012-06-27 20:44:06
@elderjameskieley: You should use quotes to specify the path to the driver executable
if the path contains spaces. According to the log, Selenium has not found the executable
and falled back to the DLL loaded into Java using JNI.
Reported by barancev
on 2012-06-28 15:29:07
I'll give it a shot, thanks for the responce
Reported by elderjameskieley
on 2012-06-28 16:02:58
It works! :D Thank you so much for all your help!!!!
Reported by elderjameskieley
on 2012-06-28 21:14:16
We've gone 24 hours without this problem after upgrading to IEDriverServer 2.24.2 (not
sure if that's what fixed it, but it seems to be gone for now). Thanks.
Reported by christoph.neuroth
on 2012-06-29 07:12:54
Reported by luke.semerau
on 2015-09-17 18:14:16
Originally reported on Google Code with ID 2946
Reported by
ruby2001
on 2011-11-28 10:19:54