First proposed by #215 and then isolated at the branch db_indexes (https://github.com/mezuro/kalibro_processor/tree/db_indexes).
We've already seen that those indexes have some impact on insertion times adding 30s to the performance test on the worst case. We need now to investigate if there is real gain on other operations which are essentially reads (in contrast to the aggregation performance test which is biased to data insertion times).
First proposed by #215 and then isolated at the branch
db_indexes(https://github.com/mezuro/kalibro_processor/tree/db_indexes).We've already seen that those indexes have some impact on insertion times adding 30s to the performance test on the worst case. We need now to investigate if there is real gain on other operations which are essentially reads (in contrast to the aggregation performance test which is biased to data insertion times).