Contributions
Testrunner returns 0 even though the project fails
I am using testrunner.bat to run my first SoapUI project. In the first step of the only test in the only test suite, an org.apache.http.conn.HttpHostConnectException occurs because of a problem in the product under test. Even though the test fails and the project fails, testrunner sets ERRORLEVEL to 0. We have scheduled the test run in Jenkins and want the job to report failure so this is of course a problem. Any idea what is wrong? Last lines of the output shown below. ... 15:58:46,860 INFO [SoapUITestCaseRunner] Finished running SoapUI testcase [Basic Object Property Access], time taken: 0ms, status: FAILED 15:58:46,865 INFO [SoapUITestCaseRunner] Project [RP-C Web Services] finished with status [FAILED] in 21226ms 15:58:46,866 ERROR [SoapUITestCaseRunner] java.lang.Exception: TestCase [Basic Object Property Access] failed without assertions 15:58:46,867 ERROR [SoapUI] An error occurred [TestCase [Basic Object Property Access] failed without assertions ], see error log for details java.lang.Exception: TestCase [Basic Object Property Access] failed without assertions at com.eviware.soapui.tools.SoapUITestCaseRunner.throwFailureException(SoapUITestCaseRunner.java:519) at com.eviware.soapui.tools.SoapUITestCaseRunner.runRunner(SoapUITestCaseRunner.java:435) at com.eviware.soapui.tools.AbstractSoapUIRunner.run(AbstractSoapUIRunner.java:202) at com.eviware.soapui.tools.AbstractSoapUIRunner.run(AbstractSoapUIRunner.java:137) at com.eviware.soapui.tools.AbstractSoapUIRunner.runFromCommandLine(AbstractSoapUIRunner.java:112) at com.eviware.soapui.tools.SoapUITestCaseRunner.main(SoapUITestCaseRunner.java:122) 15:58:46,877 WARN [Tools] Could not delete temporary directory C:\Users\SESA15~1\AppData\Local\Temp\2acace27-9289-4e23-a394-1b55a4ae789e 15:58:46,891 WARN [Tools] Could not delete temporary directory C:\Users\SESA15~1\AppData\Local\Temp\5bf817e9-fb68-45ae-a74d-5af75c5805e0 15:58:46,892 WARN [Tools] Could not delete temporary directory C:\Users\SESA15~1\AppData\Local\Temp\df6c79ad-72b3-46b3-a650-7a7fe32e5808321Views0likes0CommentsRe: Running Report Generator on a Jenkins slave
The support engineer reproduced the problem with launching Report Generator using an "Execute Windows batch command" build step. The problem is that is creates a "headless session". When the TestComplete plugin creates a session for the test, it is another session, i.e. the Report Generator and TestComplete are launched in different sessions. We successfully used a different approach to make the Report Generator and TestComplete run in the same session. We added script to our TestComplete project (see below) to execute Report Generator as the very first step. We added the following to our Python scripts: cmd = '\"C:\\Program Files (x86)\\SmartBear\\TestExecute 14\\x64\\Bin\\ReportGenerator.exe\" --desktop SE.SBO.WorkStation.exe --generate-report C:\\dev --storage-directory C:\\dev' shell = Sys.OleObject['WScript.Shell'] shell.Run(cmd, 1, False) This works well for us.2.3KViews3likes1CommentFiltering on Process in Crash Monitoring
TestComplete uses its own JIT debugger to track application crashes and capture their dumps. The debugger is triggered for all process crashes in the OS. It would be useful if we could filter by process which crashes are reported as errors in the test log. Background: Today we have trouble with TestExecute reporting crashes in mobsync.exe (Microsoft Sync Center) as errors. We have no known way to disable mobsync.exe (Offline Files is disabled but it runs anyway on every login). It is possble to work around this problem by disabling all crash monitoring (renaming tcJITHook.exe) but we prefer keeping it for relevant processes.876Views3likes1CommentRe: Running Report Generator on a Jenkins slave
That gives the same result. The console shows the full command line and hangs there without any errors reported. I tried also other variants of double-quoting (around it all and around the executable path) and in those cases I do get error messages. So, I guess Jenkins finds the Report Generator executable but for some reason it hangs trying to execute it. I have now simplified things by omitting the parameters (exe path only, with three leading and trailing ") and it hangs as well wih no errors reported (as expected).2.5KViews2likes2CommentsRe: Running Report Generator on a Jenkins slave
Not sure what you mean. We have a Jenkins service and a TestComplete service running but I suppose the TestComplete service is not involved until the TestComplete Test build step starts. If I replace the suggested command line with "C:\Windows\notepad.exe", Notepad will launch. That makes me think this is some kind of permission issue.2.5KViews0likes4CommentsRunning Report Generator on a Jenkins slave
I am trying to launch SmartBear Report Generator to be able to generate a report when TestExecute hangs on any of our Jenkins slaves, which happens intermittently. I have been told by support to run the command line below before the test starts. "C:\Program Files (x86)\SmartBear\TestExecute 14\Bin\ReportGenerator.exe" --desktop SE.SBO.WorkStation --generate-report C:\temp\DebugTrace --storage-directory C:\temp\DebugTrace I have added an "Execute Windows batch command" build step before the "TestComplete Test" build step in one of our Jenkins jobs but my build hangs when trying to execute that new build step and the Report Generator process does not start. It seems to be some kind of permission issue. If I login to the slave via Remote Desktop Connection it is possible to launch the Report Generator using the same command. Any ideas how to solve this?Solved2.6KViews0likes8CommentsRe: Running Selected Project Test Item(s) from Script
I changed the subject from "Programmatically Selecting the Project Test Items to Run" to "Running Selected Project Test Item(s) from Script" which is a more exact description of what I want to do. We have now found a solution to the problem.It is possible from script to iterate over project test items andget to the information about which script unit/routine each item is associated with. At least in Python, we can thendynamicallyimport the module(script unit)where thetest resides and execute a specificmethod (routine)that runs the test.1.9KViews1like0CommentsRunning Selected Project Test Item(s) from Script
We would like be ableto select which tests to run indifferent environments without having to login on each one of them and select the project test itemsto run.Automatic update ofproject suites in each environment is of course necessaryand we have that. We would liketo have different top items in project test items view, like "Staging"and "Production" to select the tests for each environment.Since it is not possible to launchspecific project test items from command line, we thought we couldhavespecial routines like"RunStaging" and"RunProduction" that wecall from command line,each one programmatically pointing to the top project test itemwe want to run.However, eventhough it is possible to progarmmatically access project test items,it is not possible to execute a specific item(or set certain sub items enabled before "running the project"). I know we couldcall the tests we want to run from a start routine as described above but we like the nice overview we get with theproject test items view. Anyideas?Solved1.9KViews0likes2Comments