What is the performance/complexity when we add MATCH and WHERE to a command about tile38 HOT 3 CLOSED

MATCH and WHERE add some complexity to a query. Both are linear, and are performed on each potential record. MATCH is probably a little faster than WHERE because it directly performs a comparison operation on the ID, while the WHERE needs to dig into the object fields. Both should be pretty fast in most cases, yet workflows vary so I recommend testing on your side to see which is best of you.

Also is there any performance benefit of referring the geofence as GET area test v/s defining it directly

No, both will perform the same.

from tile38.

@tidwall just had one more question regarding this. Please refer to the following query ->

SETHOOK test1 kafka://127.0.0.1:9092/bbb INTERSECTS person WHERE name 2 alice bob FENCE DETECT enter,exit OBJECT {"type":"Polygon","coordinates":[[[75.63783657767418,25.43318147102104],[75.6392084514683,25.43305944145809],........[75.63783657767418,25.43318147102104]]]}

SETHOOK test2 kafka://127.0.0.1:9092/bbb INTERSECTS person WHERE name 2 rob lob FENCE DETECT enter,exit OBJECT {"type":"Polygon","coordinates":[[[75.63783657767418,25.43318147102104],[75.6392084514683,25.43305944145809],........[75.63783657767418,25.43318147102104]]]}

as you can see the geofence is same in both the case. Does it internally treat it as 2 geofences? Or it keeps the hook on the same geofence?
If it does not treat it as same geofence i think creating a geofence first and then defining the hooks makes more sense.

from tile38.

it'll be treated as two geofences and if a devices enters the polygon it'll send two events to kafka

from tile38.