We have noticed in testing that there are some logging improvements that Synchrony Data Center requires some additional logging capability beyond its current form. This is by no means an exhaustive list, but the main points are included below:

• Database port connection check
  • If Synchrony is unable to connect to the database port, it will still start up anyway. It will not log any failures to either the Synchrony logs on the Confluence nodes or on the Synchrony Data Center nodes.
• Database bind credentials check
  • If Synchrony's startup script does not contain a valid database username/password then the startup will fail and the process will exit without a clear reason.
• Version checks of Synchrony JAR and Database JAR
  • On startup, Synchrony does not post what version of Synchrony is running, nor the database driver it is using. We would like to have this included in startup logs.
• JWT Key check
  • Synchrony requires JWT keys to startup properly. If they are not defined in environment variables or in the startup script, the startup will fail and the process will exit without a clear reason.
• Cluster node connection type not logged
  • On startup, Synchrony does not post the type of connection it is anticipating (tcpip, aws, or multicast). It also does not log if none of these are properly defined, or there is an error in the definitions.
• MD5 Mismatch
  • Synchrony will still startup if its MD5 does not match if the JAR is malformed.
• Other generic script errors
  • A malformed startup script can start up as long as the RUNJAVA details are correct for the Synchrony JAR, but will not function properly in the Confluence UI. No log output will mention missing variables or other malformed startup parameters.