GitHub user xuchuanyin reopened a pull request:
https://github.com/apache/carbondata/pull/1792 [CARBONDATA-2018][DataLoad] Optimization in reading/writing for sort temp row
Pick up the no-sort fields in the row and pack them as bytes array and skip parsing them during merge sort to reduce CPU consumption
I've tested it in my cluster and seen about 8% performance gained (74MB/s/Node -> 81MB/s/Node) in data loading. Please note that global_sort will not gain benefit from this feature since there are no sort temp file in that procedure.
Be sure to do all of the following checklist to help us incorporate
your contribution quickly and easily:
- [x] Any interfaces changed?
`Some internal used interface has been changed`
- [x] Any backward compatibility impacted?
`No`
- [x] Document update required?
`No`
- [x] Testing done
Please provide details on
- Whether new unit test cases have been added or why no new tests are required?
`No`
- How it is tested? Please attach test report.
`Tested in 3-node cluster with real business data`
- Is it a performance related change? Please attach the performance test report.
`Yes, I've tested it in my cluster and seen about 8% performance gained (74MB/s/Node -> 81MB/s/Node) in data loading.`
- Any additional information to help reviewers in testing this change.
`No`
- [x] For large changes, please consider breaking it into sub-tasks under an umbrella JIRA.
`Unrelated`
You can merge this pull request into a Git repository by running:
$ git pull
https://github.com/xuchuanyin/carbondata opt_sort_temp_serializeation
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/carbondata/pull/1792.patchTo close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #1792
----
commit de71999872761008365efc5bb943c77219479d14
Author: xuchuanyin <xuchuanyin@...>
Date: 2018-02-08T06:35:14Z
Optimization in reading/writing for sort temp row
Pick up the no-sort fields in the row and pack them as byte array and
skip parsing them during merge sort to reduce CPU consumption.
----
---
Locked
