在 AWS Glue ETL 脚本中使用自定义连接器

Question

I am working on an AWS Glue ETL script using the dynamic frame glue abstraction and writing code in python.

I created a JDBC connection resource named sap-lpr-connection in the glue data catalog and would like to use it to retrieve the connection options from the code.

As per this link (and other sources), I should be using a "custom.jdbc" connection_type to access the connection resource I created.

This is what my code looks like:

from pyspark.context import SparkContext
from awsglue.context import GlueContext

# DATABASE
database = 'sap_lpr'
table = 'bsim'

# GLUE CONTEXT
glue_context = GlueContext(SparkContext.getOrCreate())

# CONNECTION OPTIONS
connection_options = {
    "connectionName": f"{database.replace('_', '-')}-connection",
    "dbTable": table
}

# READ DATA
dyf = glue_context.create_dynamic_frame.from_options(
    connection_type="custom.jdbc",
    connection_options=connection_options
)

But when I run the code I get this error:

An error occurred while calling o81.getSource. Glue ETL Marketplace: Can not retrieve required field CONNECTOR_TYPE.

I know an alternative would be to specify a "jdbc" connection_type and pass the various connection options such as jdbc URL, username and password, but I prefer to retrieve that information from the glue connection resource I created on purpose for this.

Also, I would really like to stick to the glue_context API as opposed to the standard spark API.

Any idea what I might be doing wrong?

Answer 1

OK, it turns out that I misunderstood the type of connector I was using.

I created a connection resource in the AWS Glue Data Catalog using a "standard" connector, the JDBC one and this is not considered a custom connector type in the connection_type field, but rather a standard JDBC connection that you specify like so for example: connection_type='sqlserver' .

So if you create a connection using one of the standard connectors, such as JDBC, you have to use the .extract_from_conf() method to extract the configuration from the connection resource:

configuration = glue_context.extract_jdbc_conf(
    connection_name,
    catalog_id=None
)

connection_options = {
    "url": configuration["url"],
    "user": configuration["user"]
    "password": configuration["password"]
}