Hi everyone, can Apache Pinot supports a list of l...
# general
n
Hi everyone, can Apache Pinot supports a list of lists data type (a multi-valued column where values are also lists)? Thanks.
m
Pinot supports JSON for complex data types, you can use that?
n
yes, I can use Json but this would results in more complicated queries (we generate queries on the fly) in addition to the performance hit from using Json.
m
What’s the performance hit with JSON you are referring to?
There’s JSON indexing that performs very well.
n
From docs: It’s not performant to group by or order by a JSON field, because
JSON_EXTRACT_SCALAR
is needed to extract the values in the GROUP BY and ORDER BY clauses, which invokes the function evaluation.
I was looking for something similar to Presto’s support for ARRAY that can have ROW where ROW again can have ARRAY. Ex:
select ARRAY[ROW(ARRAY[1,2], ARRAY['a','b'])];
m
We don’t have that yet (perhaps file a GH issue with request). For performance, I’d recommend trying it out, as it depends on a lot of factors including your SLA requirement, query selectivity etc
n
thanks