• Delete your project from Rational Test Automation Server.
• Delete the configured Engineering Test Management from the Rational Test Automation Server project.

When you restart the pods, the dataset service might fail to restart due to a lock on the dataset tables.

To work around this problem, you must run the following commands before you restart the Kubernetes cluster:

1. Run the following command to get the password for the dataset service:

kubectl get secret (my-rtas}-datasets -o yaml -n {namespace}

2. Run the following command to remove the lock from the dataset table:

kubectl exec (my-rtas}-postgresql-0 -n {namespace} -i – sh -c 'PGPASSWORD=<password> psql -U datasets -d datasets -c "DROP TABLE databasechangeloglock"'

3. Run the following command to restart the Kubernetes cluster:

sudo systemctl start k3s

Note: You must substitute the following variables with the actual value in the commands:

(my-rtas} with the release name of the server that you provided during the installation of the server software.

{namespace}with the name of the namespace that you created during the installation of the server software.

<password> with the value that you obtained from Step 1.

When you have Rational Test Automation Server on Ubuntu and you run performance test assets such as VU schedule or Rate schedule on a static agent, then the test runs fail with an error message.

To work around this problem, you must add the following parameter to the helm install command during the installation of the server software:

--set networkPolicy.enabled=false

If you have already installed the server software, you can modify the helm parameter value by using the following commands:

helm upgrade {my-rtas} ./ ibm-rtas-prod -n test-system \

  --reuse-values --set networkPolicy.enabled=false

Where {my-rtas} is the name of the release that you provided during the installation of the server software.

When you back up user data of V10.1.0 and restore it on V10.1.1 in the Ubuntu environment, one of the pods does not start. This problem occurs occasionally.

For example, if the result pod does not start, to work around this problem, perform the following steps:

Note: In the following commands, <myrtas> refers to the release name of your choice and <namespace> refers to the name of the namespace that you created during the server installation process.

1. Increase the value of initialDelaySeconds in the following command to start the service:

kubectl patch sts -n <namespace> <myrtas>-results --patch '{"spec": {"template": {"spec": {"containers": [{"name": "results","readinessProbe":{"initialDelaySeconds":180}}]}}}}'

2. Get the database password by running the following command:

kubectl get secret -n <namespace> <myrtas>-results -o jsonpath="{.data.postgresql-password}" | base64 --decode

3. Delete the troublesome database lock by running the following command:

kubectl exec -n <namespace> <myrtas>-results-postgresql-0 -i -- sh -c 'PGPASSWORD=<password> psql -U results -d results -c "DROP TABLE DATABASECHANGELOGLOCK"'

4. Restart the pod by running the following command:

kubectl delete pod -n <namespace> <myrtas>-results-0

When working in multiple Rational Test Automation Server tabs concurrently, a message “You are already logged in” might sometimes be displayed. This message occurs when your session has expired and you have since logged in again, but then attempt a navigation action in a browser tab other than the one in which you logged in.

This is a known Keycloak issue. For more information, see https://issues.jboss.org/browse/KEYCLOAK-5179. To workaround this problem, close the tab and open a new one.

With 10.0, you might expect a reduction in the number of Virtual Users (VUs) supported by IBM Rational Performance Tester Agent depending upon workload characteristics. This variation is the result of the product shipping with OpenJDK™ 8 with OpenJ9 instead of IBM JRE.

OpenJDK 8 with OpenJ9 JRE exhibits a higher utilization of CPU for the workload. Up to 30% or more decrease in the supported VUs capacity may be expected for CPU-bound workloads. However, for memory-bound workloads, a much smaller, if any reduction in VUs capacity may be expected.

If you need higher performance and have access to IBM JRE 8.0.5.35 or later that contains the latest security fixes, you can switch from OpenJDK 8 with OpenJ9 to IBM JRE. For more information, see http://www.ibm.com/support/docview.wss?uid=ibm10956413.