We have seen this use-case before where DataMiner needs to offload data to systems which seem similar to the out of the box available central database feature.  (e.g. GrayLog, Kinesis Firehose, custom database schemes, ...)

How to tackle:

• Make sure DataMiner generates the offload files
• Have a driver or external software that processes these files and pushes them into the 3rd party system of choice (in your case GrayLog).

The caveat:

• In DataMiner versions prior to 10.0.10, it is not possible to have these files generated without a hack.  The 'hack' was configuring a central database with an unreachable destination.  That way the files were never deleted by DataMiner upon offload, and it allowed a driver to process these.
• From DataMiner 10.0.10 onwards there will be a clean option to generate these offload files to have a supported, future proof implementation.