跳到主要内容
版本:2.3.4

Hive

JDBC Hive Source Connector

Support Hive Version

  • Definitely supports 3.1.3 and 3.1.2, other versions need to be tested.

Support Those Engines

Spark
Flink
SeaTunnel Zeta

Key Features

supports query SQL and can achieve projection effect.

Description

Read external data source data through JDBC.

Supported DataSource Info

DatasourceSupported versionsDriverUrlMaven
HiveDifferent dependency version has different driver class.org.apache.hive.jdbc.HiveDriverjdbc:hive2://localhost:10000/defaultDownload

Database Dependency

Please download the support list corresponding to 'Maven' and copy it to the '$SEATNUNNEL_HOME/plugins/jdbc/lib/' working directory
For example Hive datasource: cp hive-jdbc-xxx.jar $SEATNUNNEL_HOME/plugins/jdbc/lib/

Data Type Mapping

Hive Data TypeSeaTunnel Data Type
BOOLEANBOOLEAN
TINYINT
SMALLINT
SHORT
INT
INTEGER
INT
BIGINTLONG
FLOATFLOAT
DOUBLE
DOUBLE PRECISION
DOUBLE
DECIMAL(x,y)
NUMERIC(x,y)
(Get the designated column's specified column size.<38)
DECIMAL(x,y)
DECIMAL(x,y)
NUMERIC(x,y)
(Get the designated column's specified column size.>38)
DECIMAL(38,18)
CHAR
VARCHAR
STRING
STRING
DATEDATE
DATETIME
TIMESTAMP
TIMESTAMP
BINARY
ARRAY
INTERVAL
MAP
STRUCT
UNIONTYPE
Not supported yet

Source Options

NameTypeRequiredDefaultDescription
urlStringYes-The URL of the JDBC connection. Refer to a case: jdbc:hive2://localhost:10000/default
driverStringYes-The jdbc class name used to connect to the remote data source,
if you use Hive the value is org.apache.hive.jdbc.HiveDriver.
userStringNo-Connection instance user name
passwordStringNo-Connection instance password
queryStringYes-Query statement
connection_check_timeout_secIntNo30The time in seconds to wait for the database operation used to validate the connection to complete
partition_columnStringNo-The column name for parallelism's partition, only support numeric type,Only support numeric type primary key, and only can config one column.
partition_lower_boundBigDecimalNo-The partition_column min value for scan, if not set SeaTunnel will query database get min value.
partition_upper_boundBigDecimalNo-The partition_column max value for scan, if not set SeaTunnel will query database get max value.
partition_numIntNojob parallelismThe number of partition count, only support positive integer. default value is job parallelism
fetch_sizeIntNo0For queries that return a large number of objects,you can configure
the row fetch size used in the query toimprove performance by
reducing the number database hits required to satisfy the selection criteria.
Zero means use jdbc default value.
common-optionsNo-Source plugin common parameters, please refer to Source Common Options for details
useKerberosBooleanNonoWhether to enable Kerberos, default is false
kerberos_principalStringNo-When use kerberos, we should set kerberos principal such as 'test_user@xxx'.
kerberos_keytab_pathStringNo-When use kerberos, we should set kerberos principal file path such as '/home/test/test_user.keytab' .
krb5_pathStringNo/etc/krb5.confWhen use kerberos, we should set krb5 path file path such as '/seatunnel/krb5.conf' or use the default path '/etc/krb5.conf '.

Tips

If partition_column is not set, it will run in single concurrency, and if partition_column is set, it will be executed in parallel according to the concurrency of tasks , When your shard read field is a large number type such as bigint( and above and the data is not evenly distributed, it is recommended to set the parallelism level to 1 to ensure that the data skew problem is resolved

Task Example

Simple:

This example queries type_bin 'table' 16 data in your test "database" in single parallel and queries all of its fields. You can also specify which fields to query for final output to the console.

# Defining the runtime environment
env {
parallelism = 2
job.mode = "BATCH"
}
source{
Jdbc {
url = "jdbc:hive2://localhost:10000/default"
driver = "org.apache.hive.jdbc.HiveDriver"
connection_check_timeout_sec = 100
query = "select * from type_bin limit 16"
}
}

transform {
# If you would like to get more information about how to configure seatunnel and see full list of transform plugins,
# please go to https://seatunnel.apache.org/docs/transform-v2/sql
}

sink {
Console {}
}

Parallel:

Read your query table in parallel with the shard field you configured and the shard data You can do this if you want to read the whole table

source {
Jdbc {
url = "jdbc:hive2://localhost:10000/default"
driver = "org.apache.hive.jdbc.HiveDriver"
connection_check_timeout_sec = 100
# Define query logic as required
query = "select * from type_bin"
# Parallel sharding reads fields
partition_column = "id"
# Number of fragments
partition_num = 10
}
}

Parallel Boundary:

It is more efficient to specify the data within the upper and lower bounds of the query It is more efficient to read your data source according to the upper and lower boundaries you configured

source {
Jdbc {
url = "jdbc:hive2://localhost:10000/default"
driver = "org.apache.hive.jdbc.HiveDriver"
connection_check_timeout_sec = 100
# Define query logic as required
query = "select * from type_bin"
partition_column = "id"
# Read start boundary
partition_lower_bound = 1
# Read end boundary
partition_upper_bound = 500
partition_num = 10
}
}