[Bugfix] Single Druid dimension spec error (part 2)

[Mogball]

issue: #3908

As a follow up to #3796, where I forgot to change the condition on the second branch, which is just an else now.

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?

[kkalyan]

This fix works for me, thanks, @Mogball.

[mistercrunch]

@fabianmenges touched this last, any comments? But there's a history of tweaks around edge cases here...

The PR seems right as we should end the if block with a catchall else: that falls back on a groupby query.

[fabianmenges]

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 queries in Druid are much faster than groupBy queries, which is why if you could use either one, your definitely want to use TopN (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 are that they can only handle a single dimension and you cannot specify the sort order.

The much better performance and the broad use case of TopN queries justify their use case. We run a very large Druid cluster (>300 machines) with very large individual datasets (>30TB) and it makes a huge difference.

[Mogball]

Wasn't aware of that. Thanks ms for the info!

[mistercrunch]

fixes #3943

[ghulands]

I just pulled in this change and found that if you don't have a dimension in the group by, it will throw an exception.

2017-11-28 20:58:09,407:ERROR:root:'dimensions'
Traceback (most recent call last):
  File "/app/.heroku/python/lib/python2.7/site-packages/superset/viz.py", line 275, in get_payload2017-11-28T20:58:09.408059+00:00 orchardapp820973[app_siri-ops_superset-druid-monitoring.web.v14]:   File "/app/.heroku/python/lib/python2.7/site-packages/superset/viz.py", line 97, in get_df
    df = self.get_df()
    self.results = self.datasource.query(query_obj)
  File "/app/.heroku/python/lib/python2.7/site-packages/superset/connectors/druid/models.py", line 1077, in query
    client=client, query_obj=query_obj, phase=2)
  File "/app/.heroku/python/lib/python2.7/site-packages/superset/connectors/druid/models.py", line 865, in get_query_str
    return self.run_query(client=client, phase=phase, **query_obj)
  File "/app/.heroku/python/lib/python2.7/site-packages/superset/connectors/druid/models.py", line 1054, in run_query
    qry['dimensions'],
KeyError: 'dimensions'