ignite-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Guillermo Ortiz <konstt2...@gmail.com>
Subject Group indices and group by query
Date Fri, 05 May 2017 17:17:53 GMT
If I have two single indices to query:

select * table a,b where a=1 and b=2

it doesn't work pretty good and I have to create a group index, how is that
possible? how does group indices work?

Similar to this:

select max(timestampField)
from myTable
group by groupField

I have a index by groupField what it means that it should be really fast to
make the group and later make a partial map to find in each node the max of
the subset of data to merge the final submax and get the final result in a
"reduce" task after sending data through network. Is it how does it work?
This query it's really slow, I don't know if I created an index by
timestampField I would get better performance, but it doesn't too much
sense to create a index with a timestampField where there aren't too row
with the same value in a table with 250M rows....
Any advice about this query?

Mime
View raw message