How to troubleshoot if an execution did not run as a SOA worker on a cloud enabled with SOA worker?

Document created by mike_aronson Employee on Jul 15, 2014
Version 1Show Document
  • View in full screen mode

You have determined by following these steps that an execution did NOT run as a SOA worker, on a cloud enabled with SOA worker.

You enabled SOA worker on your cloud, but you still do not see the worker prefix before your execution node ids.

Most likely this will occur consistently on one or more nodes in the cloud.

The steps below describe how to troubleshoot if an execution did not run as a SOA worker on a cloud enabled with SOA worker.

1. Obtain the atom cloud container logs and locate the messages in the container logs around these times:
 - when the SOA worker was enabled
 - when the execution occurred but did not run as a SOA worker

2. Look for error messages such as the following:

"Reaping abnormally terminated worker LocalWorkerJvm..."
"Discarding LocalWorkerJvm..."
"Removing worker LocalWorkerJvm..."
"Failed connecting to LocalWorkerJvm..."

3. Review the details that both precede and follow the error messages above.  They should indicate additional details with the reason for failure.
For example, a common error is "Connection Refused" error from the host and/or ports specified.    You may need to consult with your network/server admin if appropriate.

4. Verify that the port range for the SOA worker is and opened up to appropriate traffic for the nodes hosted on the cloud.

5. Make the appropriate changes to resolve the error messages that are appearing in the logs and if necessary re-start your atom cloud nodes and/or the servers hosting the cloud.

Attachments

    Outcomes