Repository: tajo Updated Branches: refs/heads/branch-0.11.1 c57956e4c -> 4c34c53c6
TAJO-1740: Update Partition Table document. (jaehwa) Project: http://git-wip-us.apache.org/repos/asf/tajo/repo Commit: http://git-wip-us.apache.org/repos/asf/tajo/commit/4c34c53c Tree: http://git-wip-us.apache.org/repos/asf/tajo/tree/4c34c53c Diff: http://git-wip-us.apache.org/repos/asf/tajo/diff/4c34c53c Branch: refs/heads/branch-0.11.1 Commit: 4c34c53c61701497b7eabc6e819a8f20d3526c5a Parents: c57956e Author: JaeHwa Jung <[email protected]> Authored: Mon Jan 18 16:11:59 2016 +0900 Committer: JaeHwa Jung <[email protected]> Committed: Mon Jan 18 16:11:59 2016 +0900 ---------------------------------------------------------------------- CHANGES | 2 + .../sphinx/partitioning/column_partitioning.rst | 213 +++++++++++++++++-- 2 files changed, 200 insertions(+), 15 deletions(-) ---------------------------------------------------------------------- http://git-wip-us.apache.org/repos/asf/tajo/blob/4c34c53c/CHANGES ---------------------------------------------------------------------- diff --git a/CHANGES b/CHANGES index 92a5a09..b695055 100644 --- a/CHANGES +++ b/CHANGES @@ -7,6 +7,8 @@ Release 0.11.1 - unreleased IMPROVEMENT + TAJO-1740: Update Partition Table document. (jaehwa) + TAJO-2053: Update description for HBase configuration. (Dongkyu Hwangbo via jaehwa) http://git-wip-us.apache.org/repos/asf/tajo/blob/4c34c53c/tajo-docs/src/main/sphinx/partitioning/column_partitioning.rst ---------------------------------------------------------------------- diff --git a/tajo-docs/src/main/sphinx/partitioning/column_partitioning.rst b/tajo-docs/src/main/sphinx/partitioning/column_partitioning.rst index 4b8d6bf..5fd44ed 100644 --- a/tajo-docs/src/main/sphinx/partitioning/column_partitioning.rst +++ b/tajo-docs/src/main/sphinx/partitioning/column_partitioning.rst @@ -11,29 +11,61 @@ How to Create a Column Partitioned Table You can create a partitioned table by using the ``PARTITION BY`` clause. For a column partitioned table, you should use the ``PARTITION BY COLUMN`` clause with partition keys. -For example, assume there is a table ``orders`` composed of the following schema. :: +For example, assume a table with the following schema. - id INT, - item_name TEXT, - price FLOAT +.. code-block:: sql + + id INT, + name TEXT, + gender char(1), + grade TEXT, + country TEXT, + city TEXT, + phone TEXT + ); -Also, assume that you want to use ``order_date TEXT`` and ``ship_date TEXT`` as the partition keys. -Then, you should create a table as follows: +If you want to make country as partitioned column, your Tajo definition would be this: .. code-block:: sql - CREATE TABLE orders ( - id INT, - item_name TEXT, - price - ) PARTITION BY COLUMN (order_date TEXT, ship_date TEXT); + CREATE TABLE student ( + id INT, + name TEXT, + gender char(1), + grade TEXT, + city TEXT, + phone TEXT + ) PARTITION BY COLUMN (country TEXT); + +Let us assume you want to use more partition columns and parquet file format. Here's an example statement to create a table: + +.. code-block:: sql + + CREATE TABLE student ( + id INT, + name TEXT, + gender char(1), + grade TEXT, + phone TEXT + ) USING PARQUET + PARTITION BY COLUMN (country TEXT, city TEXT); + +The statement above creates the student table with id, name, grade, etc. The table is also partitioned and data is stored in parquet files. + +You might have noticed that while the partitioning key columns are a part of the table DDL, theyâre only listed in the ``PARTITION BY`` clause. In Tajo, as data is written to disk, each partition of data will be automatically split out into different folders, e.g. country=USA/city=NEWYORK. During a read operation, Tajo will use the folder structure to quickly locate the right partitions and also return the partitioning columns as columns in the result set. + ================================================== -Partition Pruning on Column Partitioned Tables +Querying Partitioned Tables ================================================== -The following predicates in the ``WHERE`` clause can be used to prune unqualified column partitions without processing -during query planning phase. +If a table created using the ``PARTITION BY`` clause, a query can do partition pruning and scan only a fraction of the table relevant to the partitions specified by the query. Tajo currently does partition pruning if the partition predicates are specified in the WHERE clause. For example, if table student is partitioned on column country and column city, the following query retrieves rows in ``country=KOREA\city=SEOUL`` directory. + +.. code-block:: sql + + SELECT * FROM student WHERE country = 'KOREA' AND city = 'SEOUL'; + +The following predicates in the ``WHERE`` clause can be used to prune column partitions during query planning phase. * ``=`` * ``<>`` @@ -44,9 +76,160 @@ during query planning phase. * LIKE predicates with a leading wild-card character * IN list predicates + +================================================== +Add data to Partition Table +================================================== + +Tajo provides a very useful feature of dynamic partitioning. You don't need to use any syntax with both ``INSERT INTO ... SELECT`` and ``Create Table As Select(CTAS)`` statments for dynamic partitioning. Tajo will automatically filter the data, create directories, move filtered data to appropriate directory and create partition over it. + +For example, assume there are both ``student_source`` and ``student`` tables composed of the following schema. + +.. code-block:: sql + + CREATE TABLE student_source ( + id INT, + name TEXT, + gender char(1), + grade TEXT, + country TEXT, + city TEXT, + phone TEXT + ); + + CREATE TABLE student ( + id INT, + name TEXT, + gender char(1), + grade TEXT, + phone TEXT + ) PARTITION BY COLUMN (country TEXT, city TEXT); + + +How to INSERT dynamically to partition table +-------------------------------------------------------- + +If you want to load an entire country or an entire city in one fell swoop: + +.. code-block:: sql + + INSERT OVERWRITE INTO student + SELECT id, name, gender, grade, phone, country, city + FROM student_source; + + +How to CTAS dynamically to partition table +-------------------------------------------------------- + +when a partition table is created: + +.. code-block:: sql + + DROP TABLE if exists student; + + CREATE TABLE student ( + id INT, + name TEXT, + gender char(1), + grade TEXT, + phone TEXT + ) PARTITION BY COLUMN (country TEXT, city TEXT) + AS SELECT id, name, gender, grade, phone, country, city + FROM student_source; + + +.. note:: + + When loading data into a partition, itâs necessary to include the partition columns as the last columns in the query. The column names in the source query donât need to match the partition column names. + + ================================================== Compatibility Issues with Apache Hive⢠================================================== If partitioned tables of Hive are created as external tables in Tajo, Tajo can process the Hive partitioned tables directly. -There haven't known compatibility issues yet. \ No newline at end of file + + +How to create partition table +-------------------------------------------------------- + +If you create a partition table as follows in Tajo: + +.. code-block:: sql + + default> CREATE TABLE student ( + id INT, + name TEXT, + gender char(1), + grade TEXT, + phone TEXT + ) PARTITION BY COLUMN (country TEXT, city TEXT); + + +And then you can get table information in Hive: + +.. code-block:: sql + + hive> desc student; + OK + id int + name string + gender char(1) + grade string + phone string + country string + city string + + # Partition Information + # col_name data_type comment + + country string + city string + + +Or as you create the table in Hive: + +.. code-block:: sql + + hive > CREATE TABLE student ( + id int, + name string, + gender char(1), + grade string, + phone string + ) PARTITIONED BY (country string, city string) + ROW FORMAT DELIMITED + FIELDS TERMINATED BY '|' ; + +You will see table information in Tajo: + +.. code-block:: sql + + default> \d student; + table name: default.student + table uri: hdfs://your_hdfs_namespace/user/hive/warehouse/student + store type: TEXT + number of rows: 0 + volume: 0 B + Options: + 'text.null'='\\N' + 'transient_lastDdlTime'='1438756422' + 'text.delimiter'='|' + + schema: + id INT4 + name TEXT + gender CHAR(1) + grade TEXT + phone TEXT + + Partitions: + type:COLUMN + columns::default.student.country (TEXT), default.student.city (TEXT) + + +How to add data to partition table +-------------------------------------------------------- + +In Tajo, you can add data dynamically to partition table of Hive with both ``INSERT INTO ... SELECT`` and ``Create Table As Select (CTAS)`` statments. Tajo will automatically filter the data to HiveMetastore, create directories and move filtered data to appropriate directory on the distributed file system. +
