GitHub user sujith71955 opened a pull request:
https://github.com/apache/carbondata/pull/2913 [CARBONDATA-3090][Integration] optimizing the columnar vector filling flow by removing unwanted filter check.
## What changes were proposed in this pull request?
optimizing the columnar vector filling flow by removing unwanted filter check, currently while filling column vectors as part of vector reading flow in the carbon query, while processing the batch of rows there is a filter condition check which will validate each row whether its part of filter, since the batch already has filtered rows r , the filter check is not useful while forming the ColunVector,
in the current flow the boolean filteredRows[] is never been set once its initialized so always the values will be false.
As part of the optimization this redundant validation of filter check is been removed. This optimization can also improve the performance of queries which undergoes vectorized reading since this check for each row processing will add overhead while query processing.
## How was this patch tested?
Existing Unit test.
Be sure to do all of the following checklist to help us incorporate
your contribution quickly and easily:
- [ ] Any interfaces changed?
None
- [ ] Any backward compatibility impacted?
None
- [ ] Document update required?
None
- [ ] Testing done
Please provide details on
Existing testcases can verify the impact.
- [ ] For large changes, please consider breaking it into sub-tasks under an umbrella JIRA.
You can merge this pull request into a Git repository by running:
$ git pull
https://github.com/sujith71955/incubator-carbondata master_perf_vector
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/carbondata/pull/2913.patchTo close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #2913
----
commit 62ade40f8fd3cfeda65678828ff18a7ad64f8c3c
Author: s71955 <sujithchacko.2010@...>
Date: 2018-11-10T19:12:13Z
[CARBONDATA-3090][Integration] optimizing the columnar vector filling flow by removing unwanted filter check.
## What changes were proposed in this pull request?
optimizing the columnar vector filling flow by removing unwanted filter check, currently while filling the column vectors as part of vector reading flow in the carbon query, while processing the batch of rows there is a filter condition check which will validate each row whether its part of filter, since the batch already has filtered rows while filling the columnvector , the filter check is not useful, and even in the current flow the boolean filteredRows[] is never been set, always the values will be false.
As part of the optimization this redundant validation of filter check is been removed.
## How was this patch tested?
Existing Unit test.
----
---
Locked
