Run Matching Pipeline
A 5-step recipe with Amazon Entity Resolution. We can help you get access.
Recipe · 5 steps
Run Matching Pipeline
with Amazon Entity Resolution · Master Data Management
Create a schema mapping and matching workflow, run a job, and wait for it to finish.
- Register the schema mapping that defines the columns of the input customer records table and which columns to match on.
- Create the matching workflow that binds the new schema mapping to its input source, output destination, and rule-based resolution technique.
- Start a matching job for the newly created workflow and capture the returned job id for polling.
- Poll the job status. While the job is RUNNING or QUEUED, loop back and check again; on SUCCEEDED finish, on FAILED route to the failure step.
- Re-read the failed job so its error details are captured in the workflow outputs for diagnosis.
What you'll need to have ready
- schemaName: The name of the schema mapping to create (pattern ^[a-zA-Z_0-9-]*$).
- workflowName: The name of the matching workflow to create (pattern ^[a-zA-Z_0-9-]*$).
- description: Optional description applied to the schema mapping and workflow.
- mappedInputFields: List of SchemaInputAttribute objects (fieldName, type, optional matchKey/groupName), minimum 2.
- inputSourceARN: The Glue table ARN for the input source table.
- outputS3Path: The S3 path to which Entity Resolution writes the output table.
- output: List of OutputAttribute objects selecting columns for the output table.
- attributeMatchingModel: Either ONE_TO_ONE or MANY_TO_MANY for rule-based matching.
- rules: List of Rule objects (ruleName, matchingKeys) for RULE_MATCHING.
- roleArn: The ARN of the IAM role Entity Resolution assumes to run the workflow.
For your developer
Arazzo workflow amazon-entity-resolution:run-matching-pipeline, 5 operations: CreateSchemaMapping → CreateMatchingWorkflow → StartMatchingJob → GetMatchingJob → GetMatchingJob. Specs on APIs.io.