GitHub user shahidki31 reopened a pull request:
https://github.com/apache/carbondata/pull/1725
[CARBONDATA-1941] Documentation added for Lock Retry
The properties, default value, and description added for the lock retry.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/shahidki31/carbondata master
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/carbondata/pull/1725.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 #1725
----
commit 38038add7065a05957f84d884298c901253f3d4d
Author: mohammadshahidkhan <mohdshahidkhan1987@...>
Date: 2017-12-28T12:42:37Z
[CARBONDATA-1955] Delta DataType calculation is incorrect for long type
Problem:
In case of Long type, the delta data type is always choosing the Long type.
But it should choose the datatype based on diff (max-min) of max and min
values.
Solution:
Corrected to choose the delta data type based on max and min values.
This closes #1744
commit 1f54c47282bc201f2071bc8c9cc1be19baf0c9a1
Author: manishgupta88 <tomanishgupta18@...>
Date: 2017-12-27T17:39:58Z
[CARBONDATA-1946] Exception thrown after alter data type change operation
on dictionary exclude integer type column
Problem: After restructure change data type operation (INT to BIGINT) on
dictionary exclude INT type column if select query is triggered then exception
is thrown.
Analysis: This is happening because while retrieving the data the vector is
created for BIGINT type (size 8 bytes) which but the actual length of each data
is 4 bytes and there is length check while reading the data which is failing.
Solution: Added a new restructuredType variable in vector and assigned the
block dimension data type to it.
This closes #1732
commit e40b34b08210981cbe3bffe73cacff2577fbd8cb
Author: mohammadshahidkhan <mohdshahidkhan1987@...>
Date: 2017-06-28T13:52:45Z
[CARBONDATA-1249] Wrong order of columns in redirected csv for bad records
Problem:
Wrong order of columns in redirected csv for bad records
The RowParser rearrage the csv raw data based on the inputMapping &
outputMapping.
So the converter step does not have actual raw csv record to log or
redirect the bad record details.
Steps to repprodcue:
Create employee(Name string, age int, project string) stored by 'carbondata'
LOAD DATA LOCAL INPATH '' INTO table employee
options('BAD_RECORDS_ACTION'='REDIRECT')
Data:
Name,age,Project
Sam,27,Carbon
Ruhi,23x,Hadoop
The second record is bad record so it will be writtern to the csv file at
the bad record loation.
Expected:
Ruhi,23x,Hadoop
This closes #1116
commit 6f10c4127914c94fdd6f7af16512e0758f2bb146
Author: ravipesala <ravi.pesala@...>
Date: 2017-12-27T09:15:36Z
[CARBONDATA-1936][PARTITION] Corrected bad record and avoid double
conversion of data in Partitioning table
Currently, one time data conversion happens while loading data while
creating RDD to make sure the partitions are added with the right format. But
this approach creates an issue in case of bad record handling as the writing of
bad records not possible from RDD.
In this PR we don't convert the data in RDD but convert the data while
adding the partition information to hive.
This closes #1729
commit 7a4bd2298557ff76ee29d6c8efa37c604b548a07
Author: Manohar <manohar.crazy09@...>
Date: 2018-01-04T09:09:25Z
[CARBONDATA-1912] Handling lock issues for alter rename operation
Modified code to release all old table acquired locks in case of any
exception. So release all old table locks only in case of any exception instead
of unlocking in finally block
This closes #1734
commit bcf3ca3feda544dcbc1b5c096b98369c8a27f4e3
Author: mohammadshahidkhan <mohdshahidkhan1987@...>
Date: 2017-12-22T11:22:44Z
[CARBONDATA-1929][Validation]carbon property configuration validation +
Fixed test case
Added validation for below parameter:
carbon.timestamp.format
carbon.date.format
carbon.sort.file.write.buffer.size (minValue = 10 KB, maxValue=10MB,
defaultValue =16 KB )
carbon.sort.intermediate.files.limit (minValue = 2, maxValue=50,
defaultValue =20 )
This closes #1718
commit 837fdd2cb91780c9193efb5b37cd107c9fa36591
Author: mohammadshahidkhan <mohdshahidkhan1987@...>
Date: 2017-06-23T06:56:25Z
[CARBONDATA-1218] In case of data-load failure the
BadRecordsLogger.badRecordEntry map holding the task Status is not removing the
task Entry
Problem
For GLOBAL_SORT scope option in case of data-load failure the
BadRecordsLogger.badRecordEntry map holding the task Status is not removing the
task Entry.
Because of this the next load is getting failed even though the data being
loaded has no bad records.
Solution
The map entry must be removed after load completion either success or fail.
Refactored the Bad record logger.
This closes #1082
commit c9e58429af5f78e4001d90d6e3c90a0d48ed2bcf
Author: Jacky Li <jacky.likun@...>
Date: 2018-01-03T16:03:39Z
[CARBONDATA-1983] Remove unnecessary WriterVo object
This closes #1761
commit 08b8af7b044fa761caf58da2875935d40d4ca13b
Author: sounakr <sounakr@...>
Date: 2018-01-04T09:45:42Z
[CARBONDATA-1984][Compression Codec] Double Compression Codec Rectification.
This closes #1763
commit bbb4333f1793dae28de7a6c10cf3342204fc87b9
Author: shahidki31 <shahidki31@...>
Date: 2017-12-27T10:47:19Z
Documentation added for Lock Retry
The properties, default value, and description added for the lock retry.
----
---
