paul-rogers commented on issue #13816:
URL: https://github.com/apache/druid/issues/13816#issuecomment-1468783184
Just to put a stake in the ground, here's what _I_ would like as a user.
Declare my rollup table (i.e. "cube") in the catalog, including dimensions and
measures (what Druid call's metrics). For each measure, specify the rollup
function: `SUM(LONG)`, etc.
Then, when I do an ingest, I simply provide the values for each detail row,
and let Druid sort out which columns are dimensions and which are metrics. For
metrics, Druid works out the agg function to use. That is:
```sql
INSERT INTO myCube
SELECT
TIME_PARSE(t) AS __time,
server_name AS node,
secs * 1000 AS sum_duration_ms, ...
FROM ...
```
Here , `(__time, node)` are dimensions, `duration_ms` is a metric defined
as, say, `SUM(LONG)`. Since I already said what I wanted in the catalog
definition, I don't have to repeat myself in the ingest. This kind of thing was
done in other engines, back when cubes where hot.
This outcome is, of course, a long way off. Still, it can be helpful to
imagine where we might eventually get to.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
