New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix cardinality memory-usage considerations. #5452

Closed
wants to merge 1 commit into
base: master
from

Conversation

Projects
None yet
5 participants
@jpountz
Contributor

jpountz commented Mar 18, 2014

Default precision was computed based on the number of MULTI_BUCKET parents
instead of PER_BUCKET.

The ordinals-based execution mode was almost always used although ordinals
might have non-negligible memory usage compared to the counters.

Fix cardinality memory-usage considerations.
Default precision was computed based on the number of MULTI_BUCKET parents
instead of PER_BUCKET.

The ordinals-based execution mode was almost always used although ordinals
might have non-negligible memory usage compared to the counters.

Close #5452
@kimchy

View changes

Show outdated Hide outdated ...earch/search/aggregations/metrics/cardinality/CardinalityAggregator.java
@markharwood

This comment has been minimized.

Show comment
Hide comment
@markharwood

markharwood Mar 19, 2014

Contributor

Nice - tested on my problem dataset/query and the memory use is significantly reduced.

Contributor

markharwood commented Mar 19, 2014

Nice - tested on my problem dataset/query and the memory use is significantly reduced.

@uboness

This comment has been minimized.

Show comment
Hide comment
@uboness

uboness Mar 19, 2014

Contributor

LGTM (I can see the confusion with the MULTI_BUCKET vs PER_BUCKET... I'm open to other names that might be more descriptive... (though not that simple to find those names ;))

Contributor

uboness commented Mar 19, 2014

LGTM (I can see the confusion with the MULTI_BUCKET vs PER_BUCKET... I'm open to other names that might be more descriptive... (though not that simple to find those names ;))

@jpountz jpountz closed this in ecdcc2d Mar 20, 2014

jpountz added a commit that referenced this pull request Mar 20, 2014

Fix cardinality memory-usage considerations.
Default precision was computed based on the number of MULTI_BUCKET parents
instead of PER_BUCKET.

The ordinals-based execution mode was almost always used although ordinals
might have non-negligible memory usage compared to the counters.

Close #5452

@jpountz jpountz deleted the jpountz:fix/cardinality_memory_usage branch Mar 20, 2014

@jpountz jpountz added bug labels Mar 26, 2014

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment