Troubleshooting

4 Troubleshooting

This chapter describes where to find log files, some common error scenarios and how to troubleshoot them.

Accessing Logs

This section describes how to access logs.

Use the following commands to access the OCA logs:

kubectl get pods -n project
kubectl logs -n project <oca pod>

When the system starts, the log displays the list of capabilities cartridges loaded into this OSM instance.

Note:

You can set the OCA log level in the specification files available in the OSM cloud native toolkit if you require more detailed log information.

OSM-SCD Interaction Pattern

For every publish request initiated from Solution Designer, there is a series of interactions between the two applications. OCA logs several key messages for each of these interactions:

Publish Request

 [oracle.comms.ordermanagement.endpoints.impl.PublishOperationsEndpointImpl] - Processing asynchronously for operation id : 10050

Synchronous Response from OCA
```
publishOperations complete successful with code 202
```

Asynchronous Response from OCA

Response posted :class PublishOperationResult {
    taskId: 10316
    result: completed
    message: Deployed Cartridge: TMF_PO_B2C_Solution and version: 1.0.0.0.10050
    errors: null
}

Activate or Rollback Request

[oracle.comms.ordermanagement.endpoints.impl.PublishOperationsEndpointImpl] - Metadata refresh complete. Cartridges activated successfully

Common Error Scenarios

This section describes some of the common error scenarios you might face and how to troubleshoot them.

Product Order Stuck in "Incomplete" State

Issue: If you have not configured the emulators properly, the order will not progress to the "Completed" state.

Resolution: Configure the emulators to ensure that order processing completes correctly. For more information, refer to Installing the Emulators.

Connectivity Issues

The following error scenarios may happen if you are facing connectivity issues.

Publish Operation Returns 504 Connection Error

Issue: When you publish an initiative to OCA, it checks if the connection between SCD and OSM is functional. If you receive this error message, it signals an issue with SCD's connection to OSM.

Resolution: The possible resolution for this issue depends on the content in your initiative.

If the initiative you are publishing defines the OSM participant, then it is possible that the connection details are incorrect. If this is the case, you need to validate the endpoint URL, and the scope and authentication details configured on the Connection entity.
If you have already validated the connection details, and released the connection initiative, it is possible to see this error when you are publishing a fulfillment model. This usually indicates a transient network connection with OCA. You can do the following:
1. Retry the publish operation.
2. If the publish operation fails again, then check the status of the OCA pod in your OSM environment.
```
kubectl get pods -n project
```
  If the OCA pod is restarting, then wait until it is running and then retry publishing from Solution Designer. If the OCA pod is not running, you need to investigate and either recreate or upgrade your OSM instance to bring the pod up. If the issue persists after multiple publish attempts, you need to contact Oracle support.

Publish Operation Fails with "java.net.UnknownHostException" Error

Issue: When you publish an initiative to OCA, if the OSM hostname is not added to the host alias, the following error appears in the OSM outcome:

java.net.UnknownHostException:
                <hostname>

2025-09-16 12:53:00.165 [kafka-14] INFO o.m.w.s.impl.AsyncMessageProducer - Submit message: CloudEvent{id='ProcessTaskResponseEvent-10001-10002', source=http://workspace-manager:8080, type='workspacemanager.publishevent', data=io.cloudevents.core.data.PojoCloudEventData@fd8cc273, extensions={partitionkey=10001}} 
2025-09-16 12:53:00.182 [kafka-14] INFO o.m.w.s.impl.AsyncMessageConsumer - Received event with id: ProcessTaskResponseEvent-10001-10002, source: http://workspace-manager:8080, type: workspacemanager.publishevent, data: class ProcessTaskResponseEvent { class PublishEvent { eventType: ProcessTaskResponseEvent operationId: 10001 taskId: 10002 } response: class WsmPublishOperationResult { taskId: 10002 code: WM:09 reason: null message: java.net.UnknownHostException: kscd.osmtest3.osm.org status: null reference Error: null content: null errors: null result: failed } }

Resolution: To fix this error, in the SCD configuration file, under the workspace-manager, add a host alias section. Then, upgrade the SCD instance.

# List of host aliases.
hostAliases: []
# - ip: "127.0.0.1"
# hostnames: 
# - "kscd.osmtest3.osm.org"

Publish Operation Returns 4xx Status Code Errors

Issue: When you publish an initiative to OCA, a 4xx status code may appear in the OCA log, indicating a publish operation failure. After 30 minutes, the publish attempt fails.

If the SCD portion in the OCA secret has the wrong scope, you encounter a 403 Forbidden error.

2025-09-17 06:35:46.941 [[0x254e1d46 0x7cc449b2] WebServer socket] [INFO ] [oracle.comms.ordermanagement.util.impl.RestHttpClientUtil] - Status: 403
2025-09-17 06:35:46.941 [[0x254e1d46 0x7cc449b2] WebServer socket] [INFO ] [oracle.comms.ordermanagement.util.impl.RestHttpClientUtil] - Reason: Forbidden
2025-09-17 06:35:46.941 [[0x254e1d46 0x7cc449b2] WebServer socket] [ERROR] [oracle.comms.ordermanagement.util.impl.SCDClientImpl] - Failed to post response to http://100.76.140.68:80/scd/api/workspaceOperations/v1/publishOperationResponses/10005

If the SCD portion in the OCA secret has incorrect client details, you encounter a 401 Unauthorized error.

2025-09-17 09:02:45.640 [[0x4ed627ed 0x37348992] WebServer socket] [INFO ] [oracle.comms.ordermanagement.util.impl.RestHttpClientUtil] - Status: 401 
2025-09-17 09:02:45.640 [[0x4ed627ed 0x37348992] WebServer socket] [INFO ] [oracle.comms.ordermanagement.util.impl.RestHttpClientUtil] - Reason: Unauthorized 
2025-09-17 09:02:45.641 [[0x4ed627ed 0x37348992] WebServer socket] [ERROR] [oracle.comms.ordermanagement.util.impl.SCDClientImpl] - Failed to post response to http://100.76.140.68:80/scd/api/workspaceOperations/v1/publishOperationResponses/10008

Resolution: Update the OCA secret <project>-<test>-oca-credentials with the correct SCD client or scope details based on the issue you encountered. After that, restart the instance using the ocams option.

Runtime Issues

The following error scenarios may happen if you are facing runtime issues.

Order Creation Fails with "Product Specification Name is not a deployed entity"

Issue: When the system displays this error, it means that the product specification has not been added to the Fulfillment Model.

Resolution: Ensure that the product specification in the inbound product order exists in the Fulfillment Model included in the cartridge deployment.

Order Creation Fails with "Value Name is not a valid value for the element Element Name"

Issue: If the element "Element Name" is a lookup field, verify that "Value Name" is present in its list of values.

Resolution: Ensure that the value you want to enter for the Element Name field is present in the list of accepted values.

Troubleshooting Workspace Publish Failures

This section describes the issues you might encounter when publishing an initiative to the Workspace and how to troubleshoot them:

If an initiative publish remains in the "In Progress" state for longer than expected, it may time out. The default initiative publish duration is 30 minutes. The system then treats the attempt as a failure. Once this happens, you can reattempt the publish operation. If repeated publishing attempts fail, verify that the configuration for OCA contains the correct connection credentials and settings.
Occasionally, the system may indicate that publishing to OSM succeeded, but the overall publish failed. This happens very rarely and is caused when the Workspace Manager returns an incorrect task sequence while performing the next task. If this happens, the OSM publish operation returns a successful response, but the overall status remains stuck for the timeout duration, which is 30 minutes. After the timeout, the overall status changes to "Failed" while the OSM status remains "Successful". If this happens, retry the publish operation.