GitHub user jackylk opened a pull request:
https://github.com/apache/carbondata/pull/1987
[WIP] Add DataMap developer interface
Add developer interface for 2 types of DataMap:
1. IndexDataMap: index type of DataMap to accelerate filter query
2. OlapDataMap: olap type of DataMap to accelerate olap style query, like
SPJG query (select, predicate, join, groupby)
- [X] Any interfaces changed?
New developer interface added
- [X] Any backward compatibility impacted?
No
- [X] Document update required?
Yes
- [X] Testing done
Yes
- [X] For large changes, please consider breaking it into sub-tasks under
an umbrella JIRA.
NA
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/jackylk/incubator-carbondata
datamap-interface3
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/carbondata/pull/1987.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 #1987
----
commit 952665a8c1c52f28951463fef989333ae0e6d83e
Author: Jacky Li <jacky.likun@...>
Date: 2018-01-06T12:28:44Z
[CARBONDATA-1992] Remove partitionId in CarbonTablePath
In CarbonTablePath, there is a deprecated partition id which is always 0,
it should be removed to avoid confusion.
This closes #1765
commit 111c3821557820241d1114d87eae2f7cd017e610
Author: Jacky Li <jacky.likun@...>
Date: 2018-01-02T15:46:14Z
[CARBONDATA-1968] Add external table support
This PR adds support for creating external table with existing carbondata
files, using Hive syntax.
CREATE EXTERNAL TABLE tableName STORED BY 'carbondata' LOCATION 'path'
This closes #1749
commit 80b42ac662ebd2bc243ca91c86b035717223daf4
Author: SangeetaGulia <sangeeta.gulia@...>
Date: 2017-09-21T09:26:26Z
[CARBONDATA-1827] S3 Carbon Implementation
1.Provide support for s3 in carbondata.
2.Added S3Example to create carbon table on s3.
3.Added S3CSVExample to load carbon table using csv from s3.
This closes #1805
commit 71c2d8ca4a3212cff1eedbe78ee03e521f57fbbc
Author: Jacky Li <jacky.likun@...>
Date: 2018-01-31T16:25:31Z
[REBASE] Solve conflict after rebasing master
commit 15b4e192ee904a2e7c845ac67e0fcf1ba151a683
Author: Jacky Li <jacky.likun@...>
Date: 2018-01-30T13:24:04Z
[CARBONDATA-2099] Refactor query scan process to improve readability
Unified concepts in scan process flow:
1.QueryModel contains all parameter for scan, it is created by API in
CarbonTable. (In future, CarbonTable will be the entry point for various table
operations)
2.Use term ColumnChunk to represent one column in one blocklet, and use
ChunkIndex in reader to read specified column chunk
3.Use term ColumnPage to represent one page in one ColumnChunk
4.QueryColumn => ProjectionColumn, indicating it is for projection
This closes #1874
commit c3e99681bcd397ed33bc90e8d73b1fd33e0e60f7
Author: Jacky Li <jacky.likun@...>
Date: 2018-01-31T08:14:27Z
[CARBONDATA-2025] Unify all path construction through CarbonTablePath
static method
Refactory CarbonTablePath:
1.Remove CarbonStorePath and use CarbonTablePath only.
2.Make CarbonTablePath an utility without object creation, it can avoid
creating object before using it, thus code is cleaner and GC is less.
This closes #1768
commit e502c59a2d0b95d80db3aff04c749654254eadbe
Author: Jatin <jatin.demla@...>
Date: 2018-01-25T11:23:00Z
[CARBONDATA-2080] [S3-Implementation] Propagated hadoopConf from driver to
executor for s3 implementation in cluster mode.
Problem : hadoopconf was not getting propagated from driver to the executor
that's why load was failing to the distributed environment.
Solution: Setting the Hadoop conf in base class CarbonRDD
How to verify this PR :
Execute the load in the cluster mode It should be a success using location
s3.
This closes #1860
commit cae74a8cecea74e8899a87dcb7d12e0dec1b8069
Author: sounakr <sounakr@...>
Date: 2017-09-28T10:51:05Z
[CARBONDATA-1480]Min Max Index Example for DataMap
Datamap Example. Implementation of Min Max Index through Datamap. And Using
the Index while prunning.
This closes #1359
commit e972fd3d5cc8f392d47ca111b2d8f262edb29ac6
Author: ravipesala <ravi.pesala@...>
Date: 2017-11-15T14:18:40Z
[CARBONDATA-1544][Datamap] Datamap FineGrain implementation
Implemented interfaces for FG datamap and integrated to filterscanner to
use the pruned bitset from FG datamap.
FG Query flow as follows.
1.The user can add FG datamap to any table and implement there interfaces.
2. Any filter query which hits the table with datamap will call prune
method of FGdatamap.
3. The prune method of FGDatamap return list FineGrainBlocklet , these
blocklets contain the information of block, blocklet, page and rowids
information as well.
4. The pruned blocklets are internally wriitten to file and returns only
the block , blocklet and filepath information as part of Splits.
5. Based on the splits scanrdd schedule the tasks.
6. In filterscanner we check the datamapwriterpath from split and reNoteads
the bitset if exists. And pass this bitset as input to it.
This closes #1471
commit cd7eed66bdd7b0044953cb5bf037f6cce38c9e12
Author: xuchuanyin <xuchuanyin@...>
Date: 2018-02-08T07:39:45Z
[HotFix][CheckStyle] Fix import related checkstyle
This closes #1952
commit de92ea9a123b17d903f2d1d4662299315c792954
Author: xuchuanyin <xuchuanyin@...>
Date: 2018-02-08T06:35:14Z
[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
This closes #1792
commit 6dd8b038fc898dbf48ad30adfc870c19eb38e3d0
Author: xuchuanyin <xuchuanyin@...>
Date: 2018-02-08T06:42:39Z
[CARBONDATA-2023][DataLoad] Add size base block allocation in data loading
Carbondata assign blocks to nodes at the beginning of data loading.
Previous block allocation strategy is block number based and it will
suffer skewed data problem if the size of input files differs a lot.
We introduced a size based block allocation strategy to optimize data
loading performance in skewed data scenario.
This closes #1808
commit e5c32ac96f4cf85ef7a42f2a14c31c19418a789b
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-10T02:34:59Z
Revert "[CARBONDATA-2023][DataLoad] Add size base block allocation in data
loading"
This reverts commit 6dd8b038fc898dbf48ad30adfc870c19eb38e3d0.
commit e1c6448cdbfa8d5eab1a861485f953eea3984f1f
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-10T12:11:25Z
Revert "[CARBONDATA-2018][DataLoad] Optimization in reading/writing for
sort temp row"
This reverts commit de92ea9a123b17d903f2d1d4662299315c792954.
commit 7f5751a78c28c8a428fa62a5f82858ac65415c86
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-11T02:12:10Z
[CARBONDATA-2156] Add interface annotation
InterfaceAudience and InterfaceStability annotation should be added for
user and developer
1.InetfaceAudience can be User and Developer
2.InterfaceStability can be Stable, Evolving, Unstable
This closes #1968
commit a848ccff8aaf3c10970c61b1f85bce56478ca0ac
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-10T11:44:23Z
[CARBONDATA-1997] Add CarbonWriter SDK API
Added a new module called store-sdk, and added a CarbonWriter API, it can
be used to write Carbondata files to a specified folder, without Spark and
Hadoop dependency. User can use this API in any environment.
This closes #1967
commit 0d50f65461ae3855db66f44fa06e01174de50ccd
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-11T13:37:04Z
[CARBONDATA-2159] Remove carbon-spark dependency in store-sdk module
To make assembling JAR of store-sdk module, it should not depend on
carbon-spark module
This closes #1970
commit 937bdb867aeac5159f51de3582c4556c949bfb5c
Author: xuchuanyin <xuchuanyin@...>
Date: 2018-02-08T06:35:14Z
[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
This closes #1792
commit fd450b151cb5858504116f560da5cd7a357894e5
Author: xuchuanyin <xuchuanyin@...>
Date: 2018-02-08T06:42:39Z
[CARBONDATA-2023][DataLoad] Add size base block allocation in data loading
Carbondata assign blocks to nodes at the beginning of data loading.
Previous block allocation strategy is block number based and it will
suffer skewed data problem if the size of input files differs a lot.
We introduced a size based block allocation strategy to optimize data
loading performance in skewed data scenario.
This closes #1808
commit 0b15217106a15da8d1607eb8b7b33a29b92595a9
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-13T01:12:09Z
Support generating assembling JAR for store-sdk module
Support generating assembling JAR for store-sdk module and remove junit
dependency
This closes #1976
commit 88c0527f5421d397fb0e907e5fadd608474eadfd
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-20T03:16:53Z
[CARBONDATA-2186] Add InterfaceAudience.Internal to annotate internal
interface
This closes #1986
commit a16ec643cf48cab48f3b479a7839bd870014b463
Author: Ravindra Pesala <ravi.pesala@...>
Date: 2017-11-21T10:19:11Z
Added DataMapChooser and integrated datamapschema syntax to save datamaps
to schema
Move the expression conversion of startswith to greaterthan and lessthan
tree to carbon layer.
commit 481db5de260e9fc1d6745589997c7b78728e1e36
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-17T15:50:27Z
resolve conflict
commit bd209d9e0a018d3e88aeeff85d89eba5e287e769
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-18T17:34:15Z
fix test
commit 303380e97457b39e2e4b626d85b6b2cb58ab5924
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-20T02:53:41Z
fix test
commit 0d8c17a10455b9015ea2429275e5333d49b088d0
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-20T03:27:12Z
fix style
commit 7f8a62f9d9b64f78ad010d6bf6cde8e5f8bef0be
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-20T03:30:03Z
add comment
commit 73dabf8c23b8dd7718eac32861e0959eb13443bb
Author: Jacky Li <jacky.likun@...>
Date: 2018-02-19T19:50:23Z
datamap-interface
----
---
