...
In order to avoid the slow-access and cumbersome retrieval of data from, e.g., the IODP LIMS system, where each individual measurement is stored in its own row, the EARTHSEQUENCING database stores data aggregated on a per-Hole and per-Analysis/Datatype basis, with supporting structural meta-data, formatted as MessagePack and heavily compressed using the ZSTD algorithm. Overall this approach results in access times per hole and analysis on the order of seconds, rather than minutes as with LIMS. The actual data tables are flexible, and can be imported from traditional TSV (tab separated, preferred) or CSV files that can be exported from all major spreadsheet applications.
...
To get a general idea of the what the data tables might look like in our system, we have currently a proof-of-concept site, publicly accessible at https://paloz.marum.de. This site requires Javascript to be enabled (default on most browsers).
Please note that this will not be the final system, and furthermore does not demonstrate our workflow system that is supposed to support on-the-fly age model and depth splicing capabilities (and much more). The hooks to programmatically retrieve and search data are also not currently exposed publically, but it should be enough to give you a rough idea of how different kinds of data are imported.
...