nik9000
commented
4 years ago My pr probably could be tested with a single operation. It be good to mimick discover running against something with a couple thousand fields. I really don't want to use a full text index for this one because that isn't what I'm optimizing.
On Thu, Sep 3, 2020, 06:24 Dale McDiarmid notifications@github.com wrote:
@danielmitterdorfer https://github.com/danielmitterdorfer with the enterprise search solution this feels like its going to become more important. Recent developments in the unified highlighter https://github.com/elastic/elasticsearch/pull/61860 would of been nice to empirically measure. Do we need anything other than a single operation testing a query and highlighting on the append-no-conflicts https://github.com/elastic/rally-tracks/blob/master/pmc/challenges/default.json#L98track? Or would you prefer a separate challenge for highlighting? @dliappis https://github.com/dliappis also
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/elastic/rally-tracks/issues/9#issuecomment-686397123, or unsubscribe https://github.com/notifications/unsubscribe-auth/AABUXIS5WAINPCTUNYBBTA3SD5VGHANCNFSM4CVL4E7A .
@danielmitterdorfer with the enterprise search solution this feels like its going to become more important. Recent developments in the unified highlighter would of been nice to empirically measure. Do we need anything other than a single operation testing a query and highlighting on the append-no-conflicts track? Or would you prefer a separate challenge for highlighting? @dliappis also