Session name does not exist in in Hermes
Hi , Iam facing an issue in Linux Env . NO GUI is available An error occured [Session name 'EMS' does not exist in Hermes configuration or path to Hermes config ( /home/herrtux/.hermes )is not valid...
Forum Discussion
Hi,
We are using SOAPUI to test our integration product. All our endpoints are exposed using SOAP over JMS (TIBCO EMS).
To perform isolated tests (unit tests) we create mock-service in SOAPUI. These mock services are started simultaneously. Like many other people we frequently experience the 'Session name 'local EMS' does not exist in Hermes configuration or path to Hermes config ( C:\.hermes )is not valid !!!!' error. With this post I would like to share information and come to a solution with the SmartBear team.
Here are our observations:
- We have noticed that all needed jars are loaded (added to the classpath) more than once. In fact they are loaded again for every mock service.
This might cause the error.
- We have looked into the sources and found that the loading of the jars is not done thread safe. A flag is set once the jars have been loaded.
- When we trigger a dummy SOAP teststep before we start the rest of the mock services, the error only appears about 1 out of 5 tries, whereas if we don't add the dummy test step the hermes error appears about 9 out of ten times.
Can you confirm the above trail could be right?
Kind regards,
Harm Doedens
We are using SOAPUI to test our integration product. All our endpoints are exposed using SOAP over JMS (TIBCO EMS).
To perform isolated tests (unit tests) we create mock-service in SOAPUI. These mock services are started simultaneously. Like many other people we frequently experience the 'Session name 'local EMS' does not exist in Hermes configuration or path to Hermes config ( C:\.hermes )is not valid !!!!' error. With this post I would like to share information and come to a solution with the SmartBear team.
Here are our observations:
- We have noticed that all needed jars are loaded (added to the classpath) more than once. In fact they are loaded again for every mock service.
This might cause the error.
- We have looked into the sources and found that the loading of the jars is not done thread safe. A flag is set once the jars have been loaded.
- When we trigger a dummy SOAP teststep before we start the rest of the mock services, the error only appears about 1 out of 5 tries, whereas if we don't add the dummy test step the hermes error appears about 9 out of ten times.
Can you confirm the above trail could be right?
Kind regards,
Harm Doedens