Hi Daniel,

In my experience, VM deployment failures are quite verbosely captured in the management server logs with the default logging level that is set to DEBUG. That is, for what's relevant to Cloud Platform (orchestration/management layer). If the error is deeper like within the context of a hypervisor then you may find more information in the hypervisor logs which of course varies with the hypervisor being used.

Also, the logging levels can be modified using the log4j-cloud.xml that is located on your management server under /etc/cloud/management/. Since Cloud Platform uses standard log4j logging mechanism you can refer to log4j documentation available on the internet for more details.

The default level for both API Server and Management Server is set to DEBUG which can be increased to TRACE if required. But you shouldn't run in this mode for long. Also, changing the logging level doesn't need restart of any service, just wait for the older logs to be flushed and the newer logs will get logged using the new level.

I hope this clarifies.

Regards,
Somesh