how should lucene reindex be used?

[nguyenm100]

Hey folks, default query for lucene index is: "SELECT ?s ?p ?o ?c WHERE {{?s ?p ?o} UNION {GRAPH ?c {?s ?p ?o.}}} ORDER BY ?s"; per private volatile String reindexQuery = "SELECT ?s ?p ?o ?c WHERE {{?s ?p ?o} UNION {GRAPH ?c {?s ?p ?o.}}} ORDER BY ?s";

For larger datasets (ours is only 46M+) this is very slow often not returning. I'm guessing ordering 46M triples is not the intention. I see you can send in your own reindex query but not sure the intention beyond reindexing only a small portion of your data.

Should we consider re-writing reindex to not require order by on the query?