I am working on getting rid of many Selenium exceptions I have been getting from quite some time(Which mostly turn out to be false positives). One of the many exceptions I have been getting is the exception Selenium.SeleniumException : No SessionId provided.Due to which all the tests after that are not carried out.
The details of the exception say the following :
SetUp method failed. Selenium.SeleniumException : ERROR: ERROR: No sessionId provided. Most likely your original newBrowserSession command failed. at Selenium.HttpCommandProcessor.DoCommand(String command, String[] args) in c:\hudson\workspace\selenium-rc-trunk-win-headless\trunk\clients\dotnet\src\Core\HttpCommandProcessor.cs:line 98 at Selenium.HttpCommandProcessor.Stop() in c:\hudson\workspace\selenium-rc-trunk-win-headless\trunk\clients\dotnet\src\Core\HttpCommandProcessor.cs:line 171 at Selenium.DefaultSelenium.Stop() in c:\hudson\workspace\selenium-rc-trunk-win-headless\trunk\clients\dotnet\src\Core\DefaultSelenium.cs:line 154 at ProductionDiagnostics.TestAttributes.TearDownUseCase() in C:\BuildAgent\work\11a8046c96ca46b\ProductionDiagnostics\N开发者_StackOverflowUnitTests\TestAttributes.vb:line 121 at ProductionDiagnostics.TestAttributes.SetupUseCase() in C:\BuildAgent\work\11a8046c96ca46b\ProductionDiagnostics\NUnitTests\TestAttributes.vb:line 99
What can help me in going ahead is that why is the newbrowserSession command failing and what can be the probable causes for the same ??
Also, we recently updated the grid to 1.0.8 from 1.0.7 . Should that be a cause of trouble ??
I'm pretty sure its not a hardware problem.
You did not provide any of the code etc and but from the file path in your error syntac I can see it's Hudson who executes the tests and you will run into problem when there are several instances of Selenium RC started from Hudson as they are not thread safe then.
You need to look into using this class: ThreadSafeSeleniumSessionStorage
Update Have a look on this one: http://old.nabble.com/Problem,-Selenium-RC--sessionId--doesn't-exist--perhaps-this-session-was-already-stopped--td17633858.html
1.0.8 introduced no code that would have made a difference from 1.0.7. In all likelihood you either killed your session or Grid did it for you because it was idle for too long. Then you attempted to make a call from your client without starting a new session to handle it.
精彩评论