Is your feature request related to a problem? Please describe.
The ability to include multiple configuration files, which can have recursive includes, plus runtime resolution of variables, can be tricky to diagnose when there is unexpected behaviour/ config load errors, etc. Please add an endpoint to the HTTP server to retrieve the active config once all config includes and variable references have been processed.
This would mean it would be a lot easier to see what the final configuration looked like, and diagnose any issues. e.g. a missing include file. An environment variable reference that was misconfigured in the environment.
Describe the solution you'd like
see above - ideally a new simple GET REST endpoint.
Describe alternatives you've considered
Trying to build our own parser for the configuration. The problem with this is that it won't be the same code as Fluent Bit, so config changes/new features won't be handled correctly. As the logic to address the nesting and variable resolution exists and needs to be executed before Fluent Bit can process already exists - it makes more sense to exploit it.
Additional context
This would make diagnosing config issues easier (particularly as DryRun has its limitations)
Is your feature request related to a problem? Please describe.
The ability to include multiple configuration files, which can have recursive includes, plus runtime resolution of variables, can be tricky to diagnose when there is unexpected behaviour/ config load errors, etc. Please add an endpoint to the HTTP server to retrieve the active config once all config includes and variable references have been processed.
This would mean it would be a lot easier to see what the final configuration looked like, and diagnose any issues. e.g. a missing include file. An environment variable reference that was misconfigured in the environment.
Describe the solution you'd like
see above - ideally a new simple GET REST endpoint.
Describe alternatives you've considered
Trying to build our own parser for the configuration. The problem with this is that it won't be the same code as Fluent Bit, so config changes/new features won't be handled correctly. As the logic to address the nesting and variable resolution exists and needs to be executed before Fluent Bit can process already exists - it makes more sense to exploit it.
Additional context
This would make diagnosing config issues easier (particularly as DryRun has its limitations)