Workarounds and Limitations: Known issues in IBM Rational Quality Manager 5.0.2
The following known problems are related to Rational Quality Manager in the 5.0.2 release.
Workarounds
The following problems in this release have workarounds:
- When you export a BIRT report to Excel, you cannot open the new file in Excel
- TeamServiceRegistryException occurs when you search QM artifacts after you restore the text index
- Group By settings are not saved when you save a query for a table view
- No test case content is shown for BIRT reports that are embedded in remote dashboard widgets
- Requirement names differ between test cases and CSV spreadsheets that are downloaded from them
- When requirement collection names are updated, the old name is still shown
- Artifacts are not displayed correctly if you permanently delete the artifact template they use
- Screen capture tool does not work on macOS
- The -rebuildTextIndices command can cause the system to stop responding if work items or attachments contain long lines of text
- Connection to the mobile client fails when the server has a self-issued SSL certificate that is signed by an internal certificate authority
- Problems opening the Rational Quality Manager importer from Microsoft Excel or Microsoft Word
- The reconciliation with requirement collections fails with a timeout error
- When you use the “OAuth” XDC Authentication Type, the Quality Management ETL fails
- Text in the editor might be displayed in a language that is not installed
- Subcategory values are missing from imported process templates
- TCERs cannot be generated for a large number of test cases at one time
- Categories cannot be saved if they have the same name as a different category type in the same hierarchy
- Test case results cannot be opened if requirement links were removed from the associated test case and manual test script steps when in a read-only state
Limitations
The following problems in this release does not currently have workarounds:
- Execution variables cannot be created without a predefined value
- Online migration is not supported for upgrading some versions of the Quality Management application
Workaround summary
Problem
If you export a BIRT report to Microsoft Excel and try to open the new file in Excel, an error message about “Problems During Load” is shown. The resulting log file mentions “XML ERROR in Table”. This issue has been observed when reports include both a chart and a table.
Workaround
Open the new file by using OpenOffice Calc.
Related information
Workaround summary
Problem
After you back up (repotools -backupJFSIndexes) and restore (repotools -restoreJFSIndexes) the test index, if you search QM artifacts, an error is logged in the Rational Quality Manager log:
com.ibm.team.repository.common.transport.TeamServiceRegistryException: CRJAZ2671E The "com.ibm.rational.test.lm.sequencer.service.internal.ExecutionSequenceInformationIndexingParticipant" service failed to activate because a service that it depends on, "com.ibm.rqm.common.service.IPagedSearchService", is not registered.
If any changes were made to the Execution Schedule artifacts since the text index was backed up, they are lost.
Note: This issue impacts only Execution Schedule artifacts.
When this issue occurs, artifacts are not in the full-text index. You use that index when you search for QM artifacts from the search field in the upper-right corner.
Workaround
To retain the changes that were made to the Execution Schedule artifacts since the text index was backed up, rebuild the text index by running this command: repotools -rebuildTextIndices.
Related information
Workaround summary
Problem
This problem occurs in all the artifact table views. If you group the table view by using the Group By filter and save the query, the Group By filter is reset to Ungrouped.
Workaround
Complete these steps:
- Open the artifact table view.
- Run the query.
- From the Group By list, select a column heading to group the view as needed.
Related information
Workaround summary
Problem
This problem occurs in all the artifact table views. If you group the table view by using the Group By filter and save the query, the Group By filter is reset to Ungrouped.
Workaround
Complete these steps:
- Open the artifact table view.
- Run the query.
- From the Group By list, select a column heading to group the view as needed.
Related information
Workaround summary
Problem
In some types of BIRT reports, such as the Execution Status by Owner using TCER Count (Live) report, you can filter by test cases. However, if such a report is embedded in a dashboard widget in another application, no content is shown in the Add Test Cases window.
Workaround
Complete these steps:
- In the Add Test Cases window, show the filters section by clicking the down arrow ˇ that is just below the column headers.
- Retrieve the test cases by clicking Run.
Related information
Workaround summary
Problem
When a test case is associated with a requirement and the requirement’s name is changed, if you download the test case as a CSV spreadsheet, the spreadsheet shows the old requirement name.
Workaround
Before you download the test cases as spreadsheets, complete these steps:
- Remove the requirement link from the test case.
- Re-add the requirement link to the test case.
Related information
Workaround summary
Problem
This problem occurs when a test plan is associated with a requirement collection and the collection’s name is changed. If you open the test plan, go to the Requirement Collection Links section, and group by Requirement Collection, the old collection name is shown instead of the updated name.
Workaround
Complete these steps:
- Open the test plan in the test plan editor.
- Click Requirement Collection Links.
- Select all of the requirements and click the Reconcile Requirements in Collections icon.
Related information
Workaround summary
Problem
If you perform the Permanent Delete action on an artifact template, sections of artifacts that use the template show a Loading... message.
Workaround
Do not permanently delete an artifact template.
Related information
Workaround summary
Problem
The rich text editor in the web client provides the ability to take screen captures and attach them to artifacts. The screen capture tool is a separate Java application that requires Java Web Start to run. On macOS, if you run the downloaded JNLP file by using Java Web Start with JRE 1.6, 1.7, or 1.8, the following exception occurs: org.eclipse.swt.SWTException: Invalid thread access. This error is caused by a limitation with JNLP, which is a known Java issue.
Workaround
To insert a screen capture in the rich text editor, use the Insert Image toolbar button:- Capture an image of the screen by using an external tool.
- Save the screen capture.
- Open the Rational Quality Manager rich text editor in the web client
- Click the Insert Image toolbar button.
- In the Insert Image dialog box, click Choose File, and browse to the saved screen capture.
- Click OK to insert the screen capture in the rich text editor.
Related information
Workaround summary
Problem
In rare circumstances, when you run the repository tools -rebuildTextIndices command, the system might take a long time to respond or appear to hang. This issue occurs when text in a work item or work item attachment contains very long lines of text with no spaces, such as text that is hundreds or thousands of characters long. The system processes these lines as candidates for stack traces, which is a feature that enables you to find matching text in work items. To skip processing these lines and improve the performance of the -rebuildTextIndices command, you can set a system and server property.
Fix
The following fix is valid for these versions:- 5.0.2 iFix021 or later
- 6.0.2 iFix011 or later
- 6.0.3 iFix005 or later
- 6.0.4 and later
To process all the lines, or to change the default limitation of 2000 characters, complete the following steps.
Set the system and server property for the -rebuildTextIndices command:
1. For each server that is running an instance of the CCM or QM application, open the repotools-application.bat file.
The variable application represents the abbreviation for the application:
Change and Configuration Management (CCM): repotools-ccm.bat
Quality Management (QM): repotools-qm.bat
2. In the file, find this line:
set DEFINE=%DEFINE_INIT%
3. After the line, add these lines:
set DEFINE=%DEFINE% "-Dcom.ibm.team.repository.restrictStackTraceLineLength=false"
set DEFINE=%DEFINE% "-Dcom.ibm.team.repository.stackTraceLineLength=2000"
When this system property is set to true, any lines longer than 2000 characters will be skipped during indexing.
To adjust the default character limit, change the value of -Dcom.ibm.team.repository.stackTraceLineLength to an integer. If the value is set to 0, indexing will be skipped.
4. (Optional) For each server that is running an instance of the CCM or QM application and has encountered this issue, complete the steps based on the type of application server:
For Apache Tomcat or WebSphere Liberty:
a. In the server.startup.bat file, find this line:
set JAVA_OPTS=%JAVA_OPTS% -Djava.awt.headless=true
b. After the line, add these lines:
set JAVA_OPTS=%JAVA_OPTS% "-Dcom.ibm.team.repository.restrictStackTraceLineLength=false"
set JAVA_OPTS=%JAVA_OPTS% "-Dcom.ibm.team.repository.stackTraceLineLength=2000"
For WebSphere Application Server:
a. Use the web console to go to Application servers > server 1 > Process definition > Java Virtual Machine.
b. Add the following lines to the Generic JVM arguments settings.
-Dcom.ibm.team.repository.restrictStackTraceLineLength=false
-Dcom.ibm.team.repository.stackTraceLineLength=2000
5. Run repotools-application.bat -rebuildTextIndices.
If -Dcom.ibm.team.repository.restrictStackTraceLineLength= is set to true, lines that are longer than the specified character limit are skipped during indexing.
If debugging is enabled, the repotools-application_rebuildTextIndices log file and the console show information about the affected artifact and lines.
Example:
Indexing document belonging to artifact : 10: Stacktrace Defect , of type : com.ibm.team.workitem.WorkItem
Skipping the regex matching for line : [9/10/14 0:30:43:558 CEST] 00000101 id= ca.. of length : 2500
Skipping the regex matching for line : [9/10/14 0:31:44:557 CEST] 00000102 id= c... ; of length : 17838
Related information
- Work Item 380785
- Work Item 413255
- Repository tools command-line reference
- Repository tools command to rebuild text indexes
Workaround summary
Workaround summary
Workaround summary
Workaround summary
Workaround summary
Workaround summary
Workaround summary
Workaround summary
Workaround summary
Test case results cannot be opened if requirement links were removed from the associated test case and manual test script steps when in a read-only stateLimitation summary
Problem
Typically, you can create an execution variable without a predefined value to use in a test script step. In this scenario, the value can be specified during execution. In the 6.0 release, the display of appending the specified value to the variable during execution was incorrect. The resolution was to ensure that a placeholder is set in the test script step for the value, which requires the execution variable to have a default value. This resolution caused the regression that the execution variable cannot be created without a predefined value.
Related information
Limitation summary
Online migration is not supported for upgrading some versions of the Quality Management application
© Copyright IBM Corporation 2014, 2017