fabianmenges commented on issue #3920: [Bugfix] Single Druid dimension spec 
error (part 2)
URL: 
https://github.com/apache/incubator-superset/pull/3920#issuecomment-346372377
 
 
   > Curious, why is there a branch to specifically handle a single dimension 
(which can't be a spec), with no having_filters and order_desc = True?
   
   I can bring some light into this. 
[TopN](http://druid.io/docs/latest/querying/topnquery.html) queries in Druid 
are much faster than 
[groupby](http://druid.io/docs/latest/querying/groupbyquery.html) queries, 
which is why if you could use either one, your definitely want to use `TopN` 
(look it gets mentioned in the `groupBy` spec).
   
   `TopN` queries are much more limited than `groupBy` queries and also don't 
guarantee correctness. Their main limitations is that they can only handle a 
single dimension and you cannot specify the sort order.
   
   The much better performance and their broad use case of `TopN` queries 
justify their use case. We run a very large Druid cluster (>300 machines) with 
a lot data and it makes a huge difference.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
[email protected]


With regards,
Apache Git Services

Reply via email to