Defect fixes
You can find the defects that are fixed in this version of HCL DevOps Test Integrations and APIs (Test Integrations and APIs). You can also find the defects that were fixed in the earlier versions of Test Integrations and APIs. DevOps Test Integrations and APIs is the new name of HCL OneTest™ API from 11.0.0 onwards.
Defects fixed in DevOps Test Integrations and APIs 2024.03 (11.0.1)
The defects fixed in the current version of Test Integrations and APIs and Test Virtualization are as follows:
ID | Description |
---|---|
TP-91419 |
Previously, when you ran a virtual service on an API agent and specified the secrets collection, the tags that used secrets were not resolved against the collection. This problem is fixed. |
Defects fixed in earlier versions
You can find information about the defects fixed in each of the following versions of Test Integrations and APIs and Test Virtualization:
Defects fixed in 11.0.0
The defects fixed in Test Integrations and APIs and Test Virtualization 11.0.0 are as follows:
ID | Description |
---|---|
TP-90288 |
Previously, when you used an Oracle Database as the Result Database in your project, you might have observed errors with the SEVERE level logged in the HCL OneTest™ API-Engine logs. This might be due to the change in the version of the database drivers. This problem is fixed. |
TP-90037 |
Previously, when you resumed a paused scheduled job, HCL OneTest™ API incorrectly executed the job immediately irrespective of the scheduled time of the job. This problem is fixed. The scheduled job now resumes according to the scheduled time after it is paused. |
Defects fixed in 10.5.4
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.5.4 are as follows:
ID | Description |
---|---|
TP-90580 |
Previously, if the log level of a stub was changed when it was in the Running state in HCL® Quality Server, the change was applied without a restart of the stub. HCL® Quality Server applied the changed log level for logging into the engine log file but did not apply the change for logging into the results database. This problem is fixed. |
TP-89341 |
Previously, when you changed the scripting language in the Function action in tests, HCL OneTest™ API did not apply the correct engine when it used the test feature of the editor. This problem is fixed. |
TP-89221 |
Previously, when you used the Map action in tests, HCL OneTest™ API was not able to reference external documents that were defined in XSLT Style Sheets. This problem is fixed. |
TP-88753 |
Previously, the console log incorrectly displayed the tag name instead of the substituted response message headers in the SendReply message action in stubs. This problem is fixed. The console log now correctly displays the substituted response message headers referenced by tags in the SendReply message actions in stubs. |
TP-87618 |
Previously, if the input tag values were changed in the configuration of a running stub, the stub was stopped and restarted in a different engine. This problem is fixed. In HCL® Quality Server 10.5.4 and HCL OneTest™ API Agent 10.5.4, the changes to input tag values in the configuration of a running stub are applied without a restart of the stub. However, if you want to change any input tag values that are used within stub behaviors, you must first stop the running stub, modify the input tag values, and then restart the stub for the changed values to be applied. |
TP-88124 |
Previously, the embedded link to the results gallery of HCL® Quality Server, that is included in email reports, did not link to the specific result of that report. This problem is fixed. |
Defects fixed in 10.5.3
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.5.3 are as follows:
ID | Description |
---|---|
TP-87022 |
When you synchronized resources in an OpenApi3 document with schemas that used external references, and created an operation that was configured with the wrong json schema root, the tests that were created from the operation's MEP failed to display the JSON payloads correctly. This problem is fixed. |
Defects fixed in 10.5.2
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.5.2 are as follows:
ID | Description |
---|---|
TP-85984 |
Previously, when you tested a client application that used the Java virtualization agent to interact with a JMS stub and the stub was configured to use the Sift & pass-through stubbing mode, which resulted in a NullPointerException in the class JmsInvocationMatcherDelegateImpl. The exception occurred if a JMS request message did not specify the JMSReplyTo header attribute. This problem is fixed. |
TP-86444 |
Previously, when you attempted to configure the authentication as OAuth2 for HTTP transports, it resulted in the authentication to be applied incorrectly. This problem is fixed. |
Defects fixed in 10.5.1
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.5.1 are as follows:
ID | Description |
---|---|
TP-83559 |
Previously, when you ran different versions of stubs that used DFDL messaging schemas on multiple engines of HCL OneTest™ API agent, the engines parsed only some of the DFDL schemas. This problem is fixed. |
TP-84083 |
Previously, when invalid SWIFT payloads were used in Publish actions in tests, then HCL OneTest™ API went into the "hung" state. This problem is fixed. |
TP-84801 |
When you created a database stub from a recording with an integrated database in HCL OneTest™ API 10.5.0, the stub was empty and the following error was displayed: No driver available for virtual connection. This problem is fixed. |
TP-85240 |
If you now make changes to an identity store that is used by an HTTP transport, then the change will cause the HTTP transport to disconnect. When a test or stub that uses the HTTP transport whose identity store was changed is started subsequently, the test or stub uses the updated configuration for the HTTP transport. |
Defects fixed in 10.5.0
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.5.0 are as follows:
ID | Description |
---|---|
TP-81422 |
Previously, when you republished a stub from HCL OneTest™ API to HCL® Quality Server and the stub version was already published and running in a specific environment, the stub was reset to the first published version. This problem is fixed. Now, the republished version of the stubs correctly display their status and do not revert to their first published version. |
TP-69132 |
Previously, the error message that was displayed when there were errors due to any conversion or field expansion was lengthy and made reading of the logs difficult. This problem is fixed. The error message is now cleaned to be brief. |
TP-83872 |
Previously, when you used an IBM WebSphere MQ queue manager as the physical resource, which was configured to use the direct stubbing mode with a connection resilience, the stub failed to retry the connection if the queue manager was stopped while the stub was running. This problem is fixed. |
TP-82457 |
Previously, when installed on Windows 11, HCL OneTest™ API Agent and the Proxy might have incorrectly report the running operating system as Windows 10. This problem is fixed. The correct version of the Windows operating system is displayed now. |
Defects fixed in 10.2.3
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.2.3 are as follows:
ID | Description |
---|---|
TP-78772 |
Previously, if you configured HCL® Quality Server with Domain Level Security and configured users with roles, you might have noticed missing items in the top menu bar because the user was not assigned the appropriate role in the default domain. This problem is fixed. |
TP-80216 |
Previously, the following Subscriber field actions were not overwritten when you repaired a test in the Message Differencing Window or the Console: Asset using Function, Is Null, Length, Not Null, Regex, Schema, XSD Type. This problem is fixed. |
TP-80807 |
The Agents and Environments pages that existed in HCL® Quality Server 9.5 or earlier, are removed. If you have saved the URLs of these pages, you might need to remove them or change the links to the later versions of the Infrastructure or Environments pages of HCL® Quality Server. |
TP-81291 |
Previously, you were logged out after 30 minutes when you were on the Environments or Network and Infrastructure page in HCL® Quality Server. Although HCL® Quality Server was configured for a timeout that was set to a different period or you performed actions on these pages, you were logged out. This problem is fixed. |
TP-81799 |
Previously, when IBM MQ message data within the RFH2 header block was processed, the internal datatype attributes were incorrectly added to the Integration Tester representation of the message. This problem is fixed. |
TP-81847 |
Previously, when non-root users attempted to install HCL OneTest™ API Agent on Unix, the installation failed with an error about chmod. This problem is fixed. Now, non-root users can successfully install HCL OneTest™ API Agent on Unix. |
TP-82321 |
Previously, when you used HCL OneTest™ API 10.2.2 to record Java methods by using the Java Agent, the recording failed and an error was reported to HCL® Quality Server as: The Java Agent failed to send recording to http://<host:port>/ due to exception java.io.FileNotFoundException: http://<host:port>/ where <host:port> was the host and the port on which HCL OneTest™ API listened. This problem is fixed. |
TP-82817 |
You were not able to record HTTP messages that were larger than 1 MB via the HTTP proxy in HCL OneTest™ API 10.2.2. This problem is fixed. |
TP-82536 |
In some instances when the HTTP/TCP proxy made an onward connection by using SSL to a server that returned an HTTP 1.0 response, the client that connected via the proxy received a truncated response. This problem is fixed. |
TP-82671 |
Previously, when you used the HCL OneTest™ API JDBC driver in the same application as the Oracle ojdbc8 driver, the following error was displayed: java.lang.IllegalStateException: Timer already canceled. You can now prevent this error by specifying the following JVM argument in the application that uses the ojdbc8 driver: -Dcom.greenhat.jdbc.configure.driver.manager=false |
TP-79845 |
Previously, after you edited content in the Field Editor of expanded content and clicked OK, the Field Editor was closed after the changes that you made were applied irrespective of any error that occurred due to an incorrect change. This problem is fixed. Now, if an error occurs during an edit, you are returned to the Field Editor to correct or cancel the changes. |
TP-80321 |
Previously, JSON Schemas that were created from synchronization with a RAML document, which was located outside the project, failed to load when the RAML document was no longer available. This problem is fixed. |
TP-80791 |
Previously, when the HTTP proxy received a request that was not routed to a stub, it might have altered or added a host header in the request. This problem is fixed. |
TP-82021 |
Previously, if you deleted an environment in HCL® Quality Server and when you republished the stubs that were earlier published to the environment, which is now deleted, it resulted in HCL® Quality Server UI displaying "No resources" for the environment. This problem is fixed. |
TP-82660 |
Previously, the Java virtualization Agent failed to receive stub responses from the engine. This problem is fixed. |
Defects fixed in 10.2.2
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.2.2 are as follows:
ID | Description |
---|---|
TP-78818 |
Previously, when you ran a large number of stubs on HCL® Quality Server, you might have observed that the message log created entries for both information and warnings, which resulted in a large-sized file. Also, it was difficult to find or identify important information in the log file. This problem is fixed.Now, the log file creates entries only for the important messages. |
TP-79942 |
Previously, when you started stubs from HCL® Quality Server and found them to be connected to the results database although the stubs were configured with the logging level set to NONE, which was not the expected behavior. This problem is fixed. |
TP-80031 |
Previously, when you started a stub from HCL® Quality Server and although the stub was configured with the logging level set to NONE, the stub wrote the summary information to the results database. This incorrect behavior is fixed. |
TP-78598 |
Previously, when you synchronized resources of OpenAPI 3.0 docs in HCL OneTest™ API, the JSON schema created during synchronization incorrectly created paths to a few objects and resulted in an error such as Failed to retrieve a JSON value from the reference. This problem is fixed. |
TP-78983 |
Previously, when you attempted to run a scheduled performance test by using HCL® Quality Server, the test did not start. This impacted tests that were being run at the scheduled time or if the "Run Now" button was clicked. This problem is fixed and the tests are now able to start. |
TP-79907 |
Previously, you might have encountered issues when you integrated Groovy script jar files with HCL OneTest™ API. This problem is fixed. |
TP-79016 |
Previously, when you edited the Identity Store resources, it was possible to copy the entire content to the clipboard. This problem is fixed. Now, only the alias, type, and description are stored on the clipboard. |
TP-79133 |
Previously, the HCL OneTest™ API agent for z/OS did not provide the SQLCODE when it encountered errors during the creation and insertion of data into the simulation database. This problem is fixed. |
TP-79688 |
Previously, when you synchronized a resource in the Synchronization perspective of HCL OneTest™ API, and then attempted to promote one of the operations that was created to the root, the operation failed to be promoted to the root of the tree. This problem is fixed. |
TP-79765 |
Previously, if two instances of a virtual service resource were run simultaneously with different environments and later stopped, the virtual services displayed their status as STOPPED but the virtual services might be left running in the container. This problem is fixed. |
TP-79037 |
Previously, the Type 2 messages that were passed through by an IBM IMS Connect stub did not contain the correct security credentials. This problem is fixed. Now, HCL OneTest™ API correctly takes the security credentials from the original message for the Type 2 message that is passed through by an IBM IMS Connect stub. |
TP-79148 |
Previously, when you ran a test by using the RunTests command and when the test failed, an incorrect exit code of 0 was returned. This problem is fixed. The exit code now correctly returns 1 to indicate a failed test. |
TP-79149 |
Previously, when you ran tests or Test Suites from the UI or the command line, the summary of the test results may be incorrectly displayed as passed when the tests failed. This problem is fixed. |
TP-77518 |
Previously, when you worked with a DFDL schema and tagged values, exceptions were displayed when you viewed content in the Message Editor or when you ran a test. This problem is fixed. |
TP-58746 |
Previously, the Project Settings was slow to load in HCL OneTest™ API when it was not able to contact HCL® Quality Server. Also, when you edited the HCL® Quality Server URL in the Server Settings tab, the response from HCL® Quality Server was delayed as HCL OneTest™ API attempted to connect by using the partially entered URL. This problem is fixed. |
Defects fixed in 10.2.1
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.2.1 are as follows:
ID | Description |
---|---|
TP-77857 |
Previously, when you ran a Swagger UI request with incorrect data in the message body, an error with the HTTP status code 400 Bad Request was returned with no details about the cause of the error. This problem is fixed. The stub endpoints on HCL® Quality Server now return details about the cause of the HTTP status code 400 Bad Request and can be viewed in the Swagger UI. |
TP-77816 |
Previously, when a test suite created a report provider from a physical resource that used environment tags, the tags were not resolved. This problem is fixed. |
TP-77620 |
Previously, the Warning dialog box for the field editor did not have a Cancel button and when you clicked OK, the contents were saved and might prevent the message editor from viewing or saving the message. This problem is fixed. You can now select the Cancel action so that the validation Warning dialog box closes without saving the contents and prevents subsequent errors. |
TP-77424 |
When HCL® Quality Server is restarted, agent instances that host running virtual services are provided a longer time to relay their status and avoid duplicate instances to be created in HCL® Quality Server. The default time is set to 5 minutes. If you want to increase this time when you want to run a large number of stubs in a large number of environments, you can set the following attribute with the value for the time lag (in seconds) for the restart. You must set the attribute by appending to JVM_ARGS in the setenv.bat file that is located in the <HCL® Quality Server install dir>/bin directory.-Dstub.restart.agent.wait=<time in seconds> |
TP-77175 |
Previously, when working with an uploaded GraphQL schema in the Message Editor, when adding arguments the placeholder text "Argument Value" could cause an exception when previewing or sending content. This problem is fixed. The whitespace has been removed from this text in order to prevent the exception when displaying or sending content containing the reminder to add an argument to the node. |
TP-76392 |
Previously, in releases 10.1.1, 10.1.2, and 10.1.3, the results of tests that ran outside of suites were incorrectly recorded in the results database and the results were viewable in Results Gallery. This problem is fixed. |
TP-76081 |
Previously, when you worked with messages that had schemas such as GraphQL or GraphQL Schema applied or when you pasted GraphQL content into the Message Editor, and if query or mutation were used as part of the value of the operationName node, then the message was expanded incorrectly. This problem is fixed. |
TP-75639 |
Previously, when you installed HCL OneTest™ API Agent and HTTP Proxy on Windows 10 as a service, the operating system was displayed as Windows 8 in HCL® Quality Server. This problem is fixed. |
TP-58818 |
The following Known issue that was published in 10.0.1 is fixed. Making a configuration change to a running stub in HCL® Quality Server Environments causes it to be restarted on a new engine instance. If other stubs from the same project version are running, the new stub will attempt to listen on a different network port. This situation can disrupt routing rules. If the stubs are configured to use a fixed network port, the new instance will fail to run as the port number is already in use. Stop related stubs before making the configuration change and then start the group of stubs. |
TP-49472 |
When running test suites, failures in starting report providers are now reported in the console. |
Defects fixed in 10.2.0
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.2.0 are as follows:
ID | Description |
---|---|
TP-76682 |
Previously, when you used a File Subscriber in an Action Group, the test failed to run with an incompatible error. This problem is fixed in HCL OneTest™ API. |
TP-76598 |
Previously, when you changed the name or settings of a transport, the physical view rendering of that and other transports appeared to be truncated at times. This problem is fixed. |
TP-76115 |
Previously, when you modified or deleted a physical resource, the environment in which the physical resource was referenced was also updated unnecessarily. This problem is fixed. Now, changes such as deletion or modification to the physical resource do not affect the serialized form of the environments and HCL OneTest™ API reports the inconsistency at runtime. |
TP-76026 |
Previously, in client applications the method com.ghc.ghTester.expressions.EvalUtils.isString() in the com.ghc.ghTester JAR file became unavailable.Now, the com.ghc.ghTester JAR file once again has the method com.ghc.ghTester.expressions.EvalUtils.isString() available. |
TP-75030 |
Previously, when a JMS transport (note that this includes TIBCO EMS Domain and IBM WebSphere SlBus transports) lost its connection, HCL OneTest™ API never attempted to re-establish the connection. This problem is fixed. Note that when a JMS transport loses its connection, all existing subscriptions are terminated. |
TP-74820 |
The settings of server-based stubs that are managed in HCL® Quality Server are independent of the underlying stub definition and therefore do not change when new versions are published. The documentation is updated to clarify this behavior. |
TP-62938 |
Previously, when you promoted a field in Recording Studio and applied a filter to it, and later when you demoted the field, HCL OneTest™ API displayed the Recording Studio panel incorrectly. This problem is fixed. |
TP-76392 |
Previously, in releases 10.1.1, 10.1.2, and 10.1.3, the results of tests that ran outside of suites were incorrectly recorded in the results database and the results were viewable in Results Gallery. This problem is fixed. |
TP-76081 |
Previously, when you worked with messages that had schemas such as GraphQL or GraphQL Schema applied or when you pasted GraphQL content into the Message Editor, and if query or mutation were used as part of the value of the operationName node, then the message was expanded incorrectly. This problem is fixed. |
TP-76078 |
Previously, IBM CICS Region stubs dropped connections within 30 to 40 seconds after the connection was established. This problem is fixed. |
TP-76077 |
Previously, IBM IMS Connect stubs dropped connections within 30 to 40 seconds after the connection was established. This problem is fixed. |
TP-76037 |
Previously, when you ran a suite by using the RunTests application with the junitDir option and the results database was set to none, the run resulted in an error. This problem is fixed. |
TP-75802 |
Previously, when you worked with messages that had schemas such as GraphQL or GraphQL Schema applied, HCL OneTest™ API enclosed all arguments within quotation marks. This problem is fixed. The quotation marks are now only applied to GraphQL arguments that are either the String or ID types. |
TP-74520 |
Previously, when you worked with messages that were based on the GraphQL schema in the Message Editor, and you collapsed a GraphQL query or mutation that contained only objects with no added fields, the objects incorrectly collapsed into each other. This problem is fixed. |
TP-74489 |
Previously, when a function action that was written in ECMAScript failed, you could not identify the error easily. This problem is fixed. Now, you can identify the cause of the failure or the error in the script by the line number in the script with the error that is displayed in the console output. |
TP-74134 |
Library manager now includes providers for TIBCO EMS versions 8.4, 8.5, and 8.6. |
TP-74080 |
Previously, in the Agent.config file of Rational Integration Tester Agent, there were two sample ssl elements. Now, these are merged and there is only one ssl element in the Agent.config file. |
TP-73918 |
Previously, when you used the test feature of the scripting action, HCL OneTest™ API tried to work out what tags were being referenced and, in many scenarios, did not present the correct list. This problem is fixed. Now, you can set values for all the tags that are defined in the test rather than those that might be used by the test action. |
TP-72679 |
Previously, when a node value was reselected that was greater than 120 characters, the number of characters that were displayed in the preview text overrode the node value. This problem is fixed. Now, the number of characters in the preview text that is displayed does not affect the underlying node value. |
TP-67065 |
Previously, when you started HCL OneTest™ API, selected the Open Link option, and then clicked Cancel, HCL OneTest™ API was terminated. This problem is fixed. Now, when you start HCL OneTest™ API and select the Open Link option, and then click Cancel, you are returned to the Open Project dialog box. |
TP-49396 |
Previously, when you ran tests that use TIBCO Rendezvous messaging, the first test run failed if the TIBCO ActiveEnterprise libraries were not configured in Library Manager. This problem is fixed. |
TP-49343 |
Previously, when you attempted to send a message to a non-z/OS MQ stub that is configured for a sift and pass action, and you specified either the MQ_INPUT_AS_Q_DEF, MQ_INPUT_SHARED, or MQ_INPUT_EXCLUSIVE open option, the stub ignored the message. This problem is fixed. |
Defects fixed in 10.1.3
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.1.3 are as follows:
ID | Description |
---|---|
TP-49251 |
Previously, stubs that used an IBM WebSphere MQ transport configured to use Sift and Pass-Through did not stop when the transport was disconnected. This problem is fixed. |
TP-69433 |
Previously, when you copied and pasted a physical transport in the Physical View, a duplicate was created with the same name and you were not able to rename the copy. Also, warnings were displayed that the copy might replace the original resource. The copy action also changed the environment file. This problem is fixed. Now, when you copy and paste a physical transport, a copy is created with the same as the original transport and suffixed with <Copy {Number}>. You must rename the copy by editing the resource in the Physical View and change the settings for the copied instance. |
TP-71452 |
Previously, when you configured an encrypted password for the SSL trust store in the agent config file, an error message was displayed when the agent attempted to register with HCL® Quality Server. This problem is fixed. You can now use a password that is encrypted by using EncryptPassword.exe in the SSL trust store section of the agent config file. |
TP-71465 |
Previously, when you repaired tests with the option to overwrite the expected messages either in the Message Differences window or from the Test Lab console, only the body of the message was overwritten and the message headers were not overwritten. This problem is fixed. Now, both the message headers and body are overwritten when you select the overwrite option for expected messages. |
TP-72238 |
When you drag and drop test actions in the test editor, the CTRL key is used to determine whether the actions should be copied or moved. Previously, if the state of the CTRL key changed after you initiated the drag, the intent of the drop became ambiguous. Now, you can only perform a drop if the CTRL key is in the same state as it was when the drag was initiated. For example, if you pressed the CTRL key when you initiated a drag, you must also press the CTRL key to perform the drop as a copy action. |
TP-72449 |
Previously, when you ran the startup script to download the ISO20022 schemas, the operation terminated with an error. This problem is fixed. The URL that is embedded in the script is changed to point to the latest URL, and the script now correctly downloads the ISO20022 schemas to the default destination folder. |
TP-72523 |
Previously, when you configured an MQ queue manager for recording messages in the queue browsing mode or for direct stubbing, HCL OneTest™ API incorrectly attempted to send commands to the queue manager by using the SYSTEM.COMMAND.INPUT queue. This problem is fixed. |
TP-73092 |
Previously, the optional nodes which had nested optional container nodes that contained data were not sent in messages in certain circumstances. This problem is fixed. |
TP-39719 |
Previously, in HCL OneTest™ API, when you pasted content in an expanded node within a field editor, the node failed to update with the pasted content. This problem is fixed. |
TP-49097 |
Previously, when you selected the location to save the published results for the File publisher type in the Add Result Publisher window, you were allowed you to select a file instead of a folder or directory, which was not correct. This problem is fixed. You can now only select a folder or directory as the location to save the published result files. |
TP-49232 |
Previously, the Message Editor did not highlight the disabled fields that are not to be validated. This problem is fixed. Now, the Message Editor highlights the fields in Grey when those fields are not to be validated in the received messages. |
TP-70654 |
Previously, when you started stubs in HCL® Quality Server and the stub transports were not bound in the selected environment, the stubs might have incorrectly restarted multiple times. This problem is fixed. |
TP-72326 |
Previously, the fields that followed an unknown field in a nested Swift sequence were not marked correctly. This problem is fixed. |
TP-72527 |
Previously, when the JSON schema was applied to messages containing "any" type, the child nodes were not displayed correctly. This problem is fixed. Now, the JSON schema defaults to “Object” when the nodes in the schema use "any" type." |
TP-72612 |
Previously, in HCL OneTest™ API, when you used the switch-on-stub or switch-off-stub command in the IntegrationTesterCmd tool, the managed stub was not found if the name of the managed stub was different from the name specified in the stub definition when the stub was created. This problem is fixed. |
TP-72789 |
Previously, when you selected recorded Java method reply events and clicked the Stub icon in the Events view, the stub creation window failed to open and displayed an error. This problem is fixed. |
TP-73259 |
Previously, if you specified a host name with invalid characters for HCL OneTest™ Server in the Server Settings tab and clicked Test, no feedback was received. This problem is fixed. Now, HCL OneTest™ API displays a warning when you enter invalid characters in the host field. Also, an error is displayed when you click Test if the host name is incorrect. |
TP-73907 |
Previously, when you imported a schema such as JSON, Open API, or Swagger that contained invalid content, HCL OneTest™ API was not able to process the import and entered a “hung” state. This problem is fixed. |
Defects fixed in 10.1.2
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.1.2 are as follows:
ID | Description |
---|---|
TP-70753 |
Previously, a stub that ran on HCL® Quality Server failed to reset the restart counter when it was restarted multiple times and rendered the stub to be in a broken state. The stub might have restarted whenever HCL® Quality Server is restarted. This problem is fixed. Stubs are returned to the running state even after multiple restarts. However, you might still see the broken state logged in the logs. You can ignore the logs because the stubs run correctly after HCL® Quality Server restarts. |
TP-70739 |
Previously, when you deleted a domain from HCL® Quality Server that contained multiple versions of stubs, an inconsistent state prevented you from logging in to HCL® Quality Server when you restarted it. This problem is fixed. |
TP-70485 |
Previously, when you used a Transfer-Encoding of "chunked" and the header value was not all lower case, the HTTP proxy did not pass the response body correctly. This problem is fixed. |
TP-69611 |
Previously, when you used more than 255 characters in the tag value for stubs and published the stubs to HCL® Quality Server, such stubs were not displayed in the HCL® Quality Server Environments dashboard. This usage of Input Tags also caused the workspace to become unusable when HCL® Quality Server is restarted. This problem is fixed. You can now enter up to 4096 characters to set the value for the Input Tags for stubs. After you publish the stubs to HCL® Quality Server, the stubs display correctly in the Environments dashboard. The workspace also persists through a restart of HCL® Quality Server. |
TP-69536 |
Previously, the scenario editor in HCL® Quality Server prevented you from adding multiple instances of the same stub. Now, you can add the same stub multiple times to a scenario, and you can configure them with different settings. |
TP-69422 |
Previously, when HCL® Quality Server handled requests from multiple agents and test engines, it consumed large amounts of CPU to instantiate handlers to process the status messages. This problem is fixed. |
TP-69347 |
Previously, headers present in the RFH2 block of an IBM MQ message were sent in an unordered manner. This problem is fixed. |
TP-69327 |
Previously, the SFTP access that was used by the file transport allowed only a single logon attempt to prevent locking of the account that occurred when a wrong password was supplied. You were not able to re-established the connection by providing your password, if a network outage occurred. This problem is fixed. |
TP-68806 |
Previously, when you recorded JDBC SQL statements that contained embedded comments, the events appeared in the Recording Studio view. When you clicked the events, the SQL statement or the result set was not displayed. This problem is fixed. |
TP-68637 |
Previously, it was possible for you to add two scenarios with the same name in an Environment in HCL® Quality Server that led to confusion when identifying the scenarios. This problem is fixed. Now, when you try to save a scenario with a name that already exists, a warning is displayed and you must change the name to save the scenario. |
TP-72494 |
Previously, when you attempted to connect to a Kafka broker that was configured to use both SASL authentication and SSL, the connection failed. This problem is fixed. |
TP-71444 |
Previously, when you opened message actions that contained JSON messages, the field action configurations on JSON arrays and descendant fields that were set to values other than their default values were either removed or reverted to the default values. This problem is fixed. |
TP-70665 |
Previously, when you used the Add arrow to add a schema in the Schema Library View, the correct schema type was not selected if more than three files were added at a time. |
TP-70286 |
Previously, when you made changes to values within an operation, those changed values were not saved. This problem is fixed. |
TP-70146 |
Previously, when you used the MySQL driver from Library Manager, you were unable to connect to the Results Database. Now, you can use the supported version of MySQL V8 from Library Manager. The previous version of MySQL driver V5.1 is now labeled as deprecated in Library Manager and it is removed as a supported driver. |
TP-70058 |
Previously, when you started the RQM agent, an exception was displayed as java.nio.file.NoSuchFileException in the RQMAgent.exe log file. This problem is fixed. |
TP-69306 |
Previously, when you modified a Run Command and clicked Cancel in the Save changes dialog box, the dialog box closed incorrectly. This problem is fixed. |
TP-69137 |
Previously, data masks were not correctly applied to the recorded events after you clicked Finish in the Technique Configuration panel in the Add Data Mask wizard. This problem is fixed. |
TP-69079 |
Previously, when you added an empty CDATA[] section to an XML message, HCL OneTest™ API replaced it with a CDATA[] section that contained the null string. This problem is fixed. Now, the empty CDATA[] sections are removed from the nodes. |
TP-67138 |
Previously, recording of HTTP traffic in HCL OneTest™ API through an Istio service mesh failed on Istio V1.6 and later. This problem is fixed. |
TP-66895 |
Previously, when you changed the environment tags, any transports that used the tags were not updated with the changed values. This problem is fixed. |
TP-64932 |
Previously, when you added a property to an environment, and subsequently pulled in a change that overwrote the environment, the property that was added was seen but without an associated value. This problem is fixed. |
TP-49222 |
Previously, if the version of the configured Results database was lower than the version required, HCL OneTest™ API indicated this mismatch and if you continued with running tests, HCL OneTest™ API failed to write the results to the Results database. Now, a warning is displayed in the console that provides the cause and a resolution to fix the problem. You must update the database schema or reconfigure the project to connect to a Results database that is of a supported version. |
TP-48732 |
In HCL OneTest™ API, the JSON schema now creates roots from the $ref tag definitions. |
Defects fixed in 10.1.1
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.1.1 are as follows:
ID | Description |
---|---|
TP-70286 |
Previously, when you made changes to values within an operation, those changed values were not saved. This problem is fixed. |
TP-68924 |
HCL® Quality Server limited the number of engines that a single agent could utilize to 100. This limit is increased to 1000. The limit can be changed by editing the file:<RTCP_install_dir>/usr/servers/defaultServer/apps/RTCP.war/WEB-INF/classes/vie.server.propertiesand adding: max.engines.per.agent=<the_value_required>Then restart the server. |
TP-64814 |
The version of IBM Installation Manager that is now required to install IBM HCL OneTest™ API, HCL® Quality Server, or HCL OneTest™ API Agent is V1.9.1 or later, on all supported operating systems. Installing by using the previous versions of IBM Installation Manager is not supported. |
TP-68925 |
If you created multiple stubs to process IMS messages based on message type, transaction code, datastore, or client id, some incoming messages were routed to the wrong stub. This problem is fixed. |
TP-68927 |
You have created a test by using Send Request and Receive Reply actions to send a message to an IMS server. If the IMS system does not respond, the test will hang. This problem is fixed. |
TP-68815 |
Previously, when using HCL OneTest™ API, to generate a Jenkins project configuration, if the name of a test contained a space then quotation marks would be placed around it. It would then not be possible to run the test from Jenkins. This problem is fixed. Quotation marks are no longer added to test names when generating a Jenkins project configuration. |
TP-48644 |
When you synchronize WADL resources in HCL OneTest™ API, the Content-type is now set correctly for the HTTP headers in the synchronized operations. |
TP-48712 |
Previously, when you synchronized with WADL resources, the default selection of the resource path for property filters for the stub was incorrect. This problem is fixed. The WADL synchronization operations now have the correct default filter resource path set in HCL OneTest™ API. |
TP-48722 |
When you resynchronize WADL resources, the Schema Library now reflects the correct changes after resynchronization. |
TP-49004 |
Previously, you could not complete the copy action for the selected text in the Message Filed Editor when using the pop-up menu. This problem is fixed. The copy action is now enabled and you can complete the copy action by using the pop-up menu. |
TP-67180 |
Previously, when in the Infrastructure tab of HCL® Quality Server while viewing the Engines tab of an Agent, clicking on a stub would take you only to the last viewed Environments page. This problem is fixed. Each stub URL now returns you to the correct Environments page and displays the corresponding details for that stub. |
TP-68928 |
Previously, the recording of the HTTP traffic by using the Istio Proxy recording mode failed to start when HCL OneTest™ API ran on Linux with the java.net.SocketException:Broken pipe (Write failed) error. This problem is fixed. |
TP-49081 |
Previously, when executing a Run Command on a remote machine, if the introductory response from a Linux server was large, the action was not completed. This problem is fixed. HCL OneTest™ API can now accommodate up to 4000 characters (dependent upon encoding) of introductory text from the Linux server before the Run Command step continues successfully to run. |
TP-68125 |
Previously, when you either selected or cleared the Include optional nodes option while either applying or editing the UN EDIFACT schema, the behavior of the optional nodes in the message was incorrect. This problem is fixed. |
TP-67776 |
Previously, if an environment task was configured in the stub properties, the environment task ran incorrectly when the stub was run on a HCL OneTest™ API Agent but ran correctly when the stub was run from Rational Integration Tester. This problem is fixed. Now, an environment task now runs correctly when the stub runs on an agent or from HCL OneTest™ API. |
TP-48792 |
Previously, when a new Data mask is created in Recording Studio, the "Apply to existing events" option in the Data Mask Creation Wizard did not apply the mask to events that existed. This problem is fixed. |
Defects fixed in 10.1.0
The defects fixed in HCL OneTest™ API and HCL OneTest™ Virtualization 10.1.0 are as follows:
ID | Description |
---|---|
TP-62533 |
Previously, you could not deliver changes to a Git repository on Linux when the project was not on the same file system where HCL OneTest™ API was running. This problem is fixed. |
TP-63151 |
Previously, clicking Logout from HCL® Quality Server did not always take you to the Login page. A browser page refresh was required. Logout now always takes you to the Login page. |
TP-62977 |
Previously in HCL OneTest™ API, when you used the Recording Studio wizard to save large recorded events to a parameterized stub, in some situations, a NullPointerException was thrown before the wizard completed. This problem is fixed. |
TP-64080 |
Previously, the HTTP Proxy truncated payloads when recording at 0x00 when neither Content-Length or Transfer-Encoding were used. This problem is fixed. |
TP-65033 |
Accessing z/OS connect failed when you used a user ID whose password contained special characters. This problem is fixed. |
TP-64062 |
Previously in HCL OneTest™ API, Recording Studio events with incorrectly encoded URLs prevented tests from being created and displayed an error. Tests can now be created from these events with a clear error message displayed in the Schema Library. |
TP-64696 |
Previously, HCL OneTest™ API incorrectly saved emoji characters in tests or stubs, which resulted in errors when those tests or stubs were loaded or run. This problem is fixed. |
TP-66401 |
Previously, the proxy would delay processing an HTTP/1.0 request with no body and no content-length header. This problem is fixed. |
TP-65396 |
Previously, attempts to publish a database stub to Rational Test Control Panel failed with an exception error. This problem is fixed. |
TP-66197 |
Previously a newline character was appended to all SWIFT message types when publishing. There is now a preference on the expander and globally that allows you to control whether this newline character is appended or not. |
TP-50393 |
Previously, to use HCL OneTest™ API to record and virtualize Db2 on z/OS applications, you used the Compiled Language Interception Processor (CLIP) to intercept Db2 calls. CLIP is removed. Existing HCL OneTest™ API projects that rely on CLIP might no longer work. Rational Integration Tester now supports a new Agent for z/OS. This new agent is dependent on IBM Debugger for z/OS. |
TP-62226 |
Previously, the operation editor set the pattern to Publish/Subscribe when a Kafka or MongoDB transport was set in the binding, but failed to show the Publish/Subscribe tabs. This problem is fixed. |
TP-62478 |
The Environments, Infrastructure and Network pages of HCL® Quality Server do not detect user activity and will log the user out after 12 hours. |
TP-66529 |
When sending a MIME formatted body over HTTP, the headers in the content now override any values set in the header configuration of the editor. Values can still be overridden by setting them explicitly in the root headers of the MIME content. |
TP-61735 |
IBM WebSphere MQ transports now poll for messages in a more efficient fashion. |
TP-65902 |
Previously, when you recorded a WebSphere MQ for z/OS shared queue with queue authorization checking enabled, only messages written to the queue from the first queue manager in the queue sharing group were recorded. This problem is fixed. Now messages to the shared queue from all queue managers in the group are recorded. |
TP-46960 |
Previously, specifying anything other than the wildcard in the program and compile unit fields of the CADP profile caused an ABEND. This problem is fixed with the new Agent for z/OS. |
TP-64782 |
Previously in HCL OneTest™ API, JDBC virtualization by using MySQL failed due to incorrectly identifying the driver as com.mysql.Driver. This problem is fixed. |
TP-65063 |
Previously, when you imported a .pfx file into an identity store the following error displayed: Unable to import resources from myCert.pfx as the contents could not be identified as a JKS, PKCS#12, X.509 Certificate or Private Key. This problem is fixed and the import now proceeds correctly. |
TP-46584 |
In the Identity Store Resource, you now see an error message if you attempt to import or delete from a read-only Key Store. Previously, you could attempt to make changes to a read-only Key Store but none of these would be saved. |
TP-64660 |
Improvements were made to the storage of GIT connection preferences. |
TP-63437 |
In previous versions of HCL OneTest™ API, using a Web URL schema with a MIME multipart payload could lead to an incorrect Content-Type header being sent. This problem is fixed. |
TP-65915 |
Previously in HCL® Quality Server, when accessing the server over plain HTTP, some actions performed in the Environments page did not complete and you were returned to the Home page. This problem is fixed. |
TP-63400 |
Warning messages about Kafka request timeout settings are no longer written to the JVM console when publishing. |
TP-62875 |
Now when you deliver Git changes with HCL OneTest™ API, you can select all changes or deselect all changes. |
TP-48297 |
The default Quality Management Integration was not always accurately saved and restored. This problem is fixed. |
TP-65360 |
Previously in HCL OneTest™ API, when you started recording Istio service mesh traffic, the recording session might fail to start due to a JSON parsing exception. This problem is fixed. |
TP-62225 |
Previously in HCL® Quality Server, the start and stop stub API endpoints returned an Internal Server Error (status code 500) when a Content-Type header was not sent in the request. This problem is fixed. |
TP-61867 |
Previously, when there were no test results in Jenkins to review, the following incorrect message was displayed: No results link from Ant task found in build log. Now, the correct message is displayed: No results link from Integration Tester task found in build log. |
TP-59765 |
With valid configuration settings on a physical Email server resource, a Publish action of a test that used that transport worked correctly, however the Test Transport button failed to connect to the server. This problem is fixed. |
TP-61802 |
Previously, large projects publishing to HCL® Quality Server could take several minutes. Though the stubs published to HCL® Quality Server successfully, you could see a message "Stubs were not published due to the following error: HCL OneTest™ API could not connect to the specified HCL® Quality Server." This problem is fixed. |
TP-58766 |
In the Message Differences Window, when you use either Overwrite Expected Message or Overwrite Expected Field, null values from the actual message are now correctly assigned isNull action types and are successfully overwritten into the expected message. |
TP-63198 |
Previously, in HCL® Quality Server, when you clicked Logout, a browser refresh was required to take you to the Login page. Now when you click Logout, you are taken back to the Login page. |
TP-64275 |
When recording database queries that contain subqueries, the rows in one or more subquery tables are not populated in the virtual database if the subquery is in the WHERE clause and uses an IN, ANY or ALL expression. This problem is fixed. |
TP-64968 |
Previously, an error was shown in the console when you ran a test by using RunTests with a blank password. This problem was fixed. |
TP-64310 |
Previously, when you entered a parameterFile as an argument, you were shown a warning. This warning no longer occurs as long as the parameterFile argument is used outside of the file that is being referenced by this argument. |
TP-64910 |
Previously, query strings were shown in the Resource Path of the Stub panel when this operation was created from traffic recorded in Recording Studio. To simplify the creation of a stub from the recording traffic, the query strings are now removed from the Resource Path shown in the Stub panel. |
TP-64097 |
Previously, when you clicked the Analyze Schema button in the Schema Library, HCL OneTest™ API would hang. This problem is fixed. Now when you click the Analyze Schema button, results display as expected. |
TP-62359 |
When you use MySQL while running a performance test from HCL OneTest™ API, an error can occur for the GROUP BY clause. This error is caused when ONLY_FULL_GROUP_BY is set in the sql_mode GLOBAL setting. To work around this problem, see Troubleshooting: Running a performance test with a MySQL database. |
TP-66661 |
Previously, when you used the Environments, Infrastructure or Networks dashboards of HCL® Quality Server the console.log included a warning that a file was not found even though the dashboard was displayed correctly. This problem is fixed so these incorrect additions to the console.log file no longer occur. |
TP-64669 |
Previously, when you used the Update Project or Undo changes action in HCL OneTest™ API to refresh the local repository with the changes from the remote repository, any editors that were open were not refreshed with the changes. This problem is fixed. |