GitHub user ravipesala opened a pull request:
https://github.com/apache/carbondata/pull/2324
[CARBONDATA-2496] Changed to hadoop bloom implementation and added compress
option to compress bloom on disk.
This PR removes the guava bloom and adds the hadoop bloom. And also added
the compress bloom option to compress bloom on disk and in memory as well.
The user can use `bloom_compress` property to enable/disable compression.
By default, it is enabled.
Please check the performance of bloom
Loaded 100 million data with bloom datamap on a column with a cardinality
of 5 million with 'BLOOM_SIZE'='5000000', 'bloom_fpp'='0.001' .
Guava
-----------------------------
DataMap Size : 233.6 MB
Fisrt Read : 4.981 sec
Second Read: 0.541 sec
Hadoop
-----------------------------------
DataMap Size : 224.7 MB
Fisrt Read : 4.829 sec
Second Read: 0.555 sec
Haoop with compression
------------------------------------
DataMap Size : 98.8 MB
Fisrt Read : 4.984 sec
Second Read: 0.621 sec
As per the above readings compressed Hadoop implementation gives optimal
performance in terms of space and read.Thats why compress is made as default.
Here RoaringBitMap is used internally for compressing the bloom so it s not
only space efficient on disk and also efficient in memory.
Be sure to do all of the following checklist to help us incorporate
your contribution quickly and easily:
- [ ] Any interfaces changed?
- [ ] Any backward compatibility impacted?
- [ ] Document update required?
- [ ] Testing done
Please provide details on
- Whether new unit test cases have been added or why no new tests
are required?
- How it is tested? Please attach test report.
- Is it a performance related change? Please attach the performance
test report.
- Any additional information to help reviewers in testing this
change.
- [ ] 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/ravipesala/incubator-carbondata bloom-change1
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/carbondata/pull/2324.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #2324
----
commit 5bf9726bd5f5703bb8ef51abf6a6c75bc5423586
Author: ravipesala <ravi.pesala@...>
Date: 2018-05-20T16:22:57Z
Changed to hadoop bloom implementation and added compress option to
compress bloom on disk.
----
---
