At present, we don't employ carriage returns, which makes it non-trivial to divide up the 9 main JSON files for indexing by ElasticSearch. If we put each record on its own line, then we could divide up files and index them without having to even look at the content, requiring considerably less horsepower than would otherwise be necessary.
To be clear, this would be an alternative to generating literally millions of individual JSON files.
At present, we don't employ carriage returns, which makes it non-trivial to divide up the 9 main JSON files for indexing by ElasticSearch. If we put each record on its own line, then we could divide up files and index them without having to even look at the content, requiring considerably less horsepower than would otherwise be necessary.
To be clear, this would be an alternative to generating literally millions of individual JSON files.