AutoGineer
Generated banner for Ingesting Into Data Lake
Generated logo for Ingesting Into Data Lake

Ingesting Into Data Lake

AWSOfficialPublished by AWS

  • Skill
  • Workflow
  • Data and AI
  • Apache-2.0
  • Low risk
Download ZIP

From the author

Import data into the AWS data lake from S3 files, local uploads, JDBC databases (Oracle, SQL Server, PostgreSQL, MySQL, RDS, Aurora), Amazon Redshift, Snowflake, BigQuery, DynamoDB, or existing Glue catalog tables (migration). Default target is S3 Tables; standard Iceberg on a general purpose bucket is supported where S3 Tables is not adopted. Handles one-time loads, recurring pipelines, migrations. Triggers on: import data, load data, ingest, sync database, migrate table, move data to AWS, set up pipeline, ETL, pull from Snowflake, query BigQuery into S3, export DynamoDB, CTAS, convert to Iceberg. Do NOT use for setting up or troubleshooting Glue connections (use connecting-to-data-source), creating empty tables (use creating-data-lake-table), running queries (use querying-data-lake), finding tables by fuzzy name (use finding-data-lake-assets), catalog audit (use exploring-data-catalog), or SaaS platforms like Salesforce, ServiceNow, SAP, MongoDB, Kafka.
— the description by AWS

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

This skill provides a structured workflow for importing data from various sources—including JDBC databases, cloud warehouses like Snowflake and BigQuery, DynamoDB, and S3 files—into an AWS data lake. It prioritizes modern S3 Tables or standard Iceberg formats, ensuring data is queryable and well-structured. The skill guides users through connection verification, target configuration, and validation, while delegating complex setup tasks to specialized sub-skills.

What it does

  • Imports data from JDBC databases, cloud warehouses, and S3 into a data lake
  • Supports one-time loads, recurring ETL pipelines, and catalog migrations
  • Configures target tables using S3 Tables or standard Iceberg formats
  • Validates data integrity via row counts, null checks, and sample verification
  • Automates scheduling for recurring ingestion tasks using Glue triggers
  • Provides clear delegation paths for connection setup and table creation

When to use it

  • When you need to move data from an external database or cloud service into your AWS data lake
  • When setting up a new recurring ETL pipeline for data synchronization
  • When migrating existing Hive-style tables to modern Iceberg formats
  • When you need to perform a one-time bulk load of files from S3 or local storage

Advantages

  • Standardizes ingestion patterns across multiple source types
  • Ensures data quality through built-in validation steps
  • Promotes modern data lake practices by defaulting to S3 Tables and Iceberg
  • Reduces configuration errors by enforcing connection verification before execution

Usage examples

“Import the customers table from our Oracle database into the data lake.”
The agent verifies the existing JDBC connection, confirms the target S3 Tables configuration, and initiates a Glue ETL job to migrate the data.
“Set up a recurring pipeline to pull data from BigQuery into S3.”
The agent configures a Glue job with the necessary BigQuery connection, sets up the target Iceberg table, and creates a Glue trigger for scheduled execution.
“Convert our existing Hive tables to Iceberg.”
The agent performs a catalog migration, updating the table metadata to Iceberg format while maintaining data accessibility.
“Load these local CSV files into the data lake.”
The agent guides the user through the local upload process and maps the data into a queryable S3 Tables format.

Install

  1. Download the skill as a ZIP:

    https://mp-staging.autogineer.com/agent-skills/aws/ingesting-into-data-lake/download

  2. Unzip it into the cross-client skills folder of your project:

    mkdir -p .agents/skills && unzip ingesting-into-data-lake.zip -d .agents/skills

Documentation by AWS

Ingest into Data Lake

Move data from a source into a queryable table in the data lake. This skill assumes the source connection (if one is needed) already exists. For Glue connection setup or troubleshooting, delegate to connecting-to-data-source.

Philosophy

Default to S3 Tables unless the environment says otherwise. S3 Tables is the recommended target for new data lake work. If the user's catalog inventory shows they haven't adopted S3 Tables, recommend standard Iceberg on their existing general-purpose bucket instead of forcing them to change posture.

Common Tasks

You MUST execute commands using AWS MCP server tools when connected -- they provide validation, sandboxed execution, and audit logging. Fall back to AWS CLI only if MCP is unavailable. You MUST explain each step before executing.

Workflow

1. Verify Dependencies and Context

  • You MUST check whether AWS MCP tools or AWS CLI are available and inform the user if missing

  • You MUST confirm target AWS region and verify credentials with aws sts get-caller-identity

  • For SageMaker Unified Studio project roles, note that target tables and connections may be scoped to the project. See the caller ARN detection pattern in querying-data-lake.

2. Classify the Source
User says...Source typeReference
"upload my file", "local CSV", "move to S3"Local filelocal-upload.md
"load from S3", "import CSV/JSON/Parquet from s3://"S3 filess3-files.md
"import from Oracle/Postgres/MySQL/SQL Server/Redshift/RDS/Aurora"JDBCjdbc-ingest.md
"pull from Snowflake", "Snowflake table to S3"Snowflakesnowflake-ingest.md
"import from BigQuery", "GCP analytics to S3"BigQuerybigquery-ingest.md
"export DynamoDB", "DynamoDB to data lake"DynamoDBdynamodb-ingest.md
"migrate Glue table", "convert Hive to Iceberg"Catalog migrationcatalog-migration.md

If the user names Salesforce, ServiceNow, SAP, MongoDB, Kafka, or another SaaS/streaming source, decline -- these are not supported in this release.

If the source table is referenced by a fuzzy or business name ("migrate our orders table", "pull from the sales warehouse"), delegate to finding-data-lake-assets to resolve before proceeding.

3. Confirm Connection Exists (if applicable)

For JDBC, Snowflake, and BigQuery sources, a Glue connection is required. Check:

aws glue get-connection --name <CONNECTION_NAME> --region <REGION>

If the connection does not exist, stop and delegate to connecting-to-data-source to create and test it. Do not proceed with ingest until the connection is verified.

Local files, S3 files, DynamoDB, and catalog migration do not need a Glue connection.

4. Clarify the Target

You MUST ask the user (or suggest based on catalog inventory) before creating or writing to any table:

  • Database/namespace: Does a specific target database exist? Or should one be created?

  • Table: Existing table (append/merge) or new table (delegate to creating-data-lake-table)?

  • Format: S3 Tables (default), standard Iceberg, or raw Parquet?

Inventory-aware defaults:

If you have already run exploring-data-catalog or can quickly check, use what exists:

  • Account has an s3tablescatalog federated catalog and active table buckets: recommend S3 Tables

  • Account has general-purpose buckets with Iceberg tables and no S3 Tables usage: recommend standard Iceberg on their existing bucket

  • Account uses Parquet/ORC on S3 without Iceberg metadata: ask whether to adopt Iceberg now (recommend yes) or continue with raw files

Do not force S3 Tables on customers who haven't adopted it. See iceberg-catalog-config-and-usage.md.

Delegations from this step:

  • Target table doesn't exist -> creating-data-lake-table

  • Target database named by fuzzy term -> finding-data-lake-assets

  • User doesn't know what exists -> exploring-data-catalog

5. Execute Source Workflow

Read the source-specific reference and follow its phases. Each is self-contained with job templates, gotchas, and troubleshooting:

  • Local / S3 / JDBC / Snowflake / BigQuery / DynamoDB / catalog migration -- one reference per source

Common Glue 5.1 or higher job configuration and PySpark templates are shared in glue-job-config.md and glue-job-scripts.md.

6. Validate

Run all three, do not skip:

  1. Row count matches expected (source vs target)

  2. Null check on critical columns

  3. Spot-check 3-5 sample rows

See data-quality-validation.md.

7. Schedule (if recurring)

For recurring pipelines, create a Glue Trigger with a cron schedule. See testing-and-scheduling.md. Simple single-step pipelines use Glue Triggers; multi-step with branching uses MWAA.

Argument Routing

  • S3 path only: Infer one-time load, start Step 2 with S3 files

  • Connection name: Start Step 3 with the named connection

  • Table name: Start Step 4, ask whether this is source or target

  • --target flag: Pre-fill the target format in Step 4

  • No args: Walk through interactively

Gotchas

  • S3 Tables requires Glue 5.1 or higher and --datalake-formats iceberg job argument

  • All spark.sql.catalog.* config MUST go in --conf job arguments, never in spark.conf.set(). Glue 5.x throws AnalysisException: Cannot modify the value of a static config otherwise. See iceberg-catalog-config-and-usage.md for correct catalog configs.

  • The warehouse parameter is required in S3 Tables catalog config. Without it Spark fails with "Cannot derive default warehouse location".

  • Table and column names in S3 Tables MUST be all lowercase

  • overwritePartitions() only replaces partitions present in the DataFrame -- for full refresh with deletes, use createOrReplace()

  • Standard Iceberg targets MUST include a LOCATION clause; S3 Tables MUST NOT

  • DynamoDB does not need a Glue connection -- do not attempt to create one

  • Connection failures during ingest delegate back to connecting-to-data-source; do not debug network/credentials in this skill

  • For target tables in SageMaker Unified Studio projects, ensure the project role has write access to the target namespace before the Glue job runs

Troubleshooting

ErrorLikely causeAction
Access Denied on S3Missing IAM permissionsCheck Glue role has s3:GetObject, s3:PutObject
Access Denied on S3 TablesMissing s3tables:* permissionsAdd S3 Tables inline policy to Glue role
CTAS timeoutDataset too large for AthenaSwitch to Glue ETL or batch with WHERE filters
JDBC connection timeout/auth failureConnection-level issueDelegate to connecting-to-data-source
Throughput exceeded (DynamoDB)Read percent too highLower read.percent or use native export

See error-handling.md for the full catalog.

References

Source-specific
Cross-cutting
Migration-specific
JDBC-specific

Files

references/schema-evolution.md

Schema Evolution and Nested Structure Handling Reference

This document describes expected approaches for handling schema evolution and nested JSON/struct data during imports.

Schema Evolution

What is Schema Evolution?

Schema evolution occurs when source data has columns that don't exist in the target table. This is common when:

  • Source data schema changes over time (new fields added)

  • Importing from multiple sources with different schemas

  • Business requirements evolve and new data points are captured

Types of Schema Changes
Change TypeExampleHandling
New columnsSource has phone_number, table doesn'tALTER TABLE ADD COLUMNS
Missing columnsTable has country, source doesn'tUse NULL or default value
Type changesSource price is STRING, was INTType conflict resolution (see type-transformations.md)
Column renameSource has customer_name, table has nameManual mapping or user decision

Schema Evolution Workflow

1. Detect Schema Differences
# Get current table schema from Glue Catalog
import boto3
glue = boto3.client('glue')

response = glue.get_table(
    DatabaseName='my_database',
    Name='my_table'
)

existing_columns = {col['Name']: col['Type'] for col in response['Table']['StorageDescriptor']['Columns']}

# Compare with source schema
source_columns = {'customer_id': 'int', 'name': 'string', 'email': 'string', 'phone': 'string'}  # Inferred

new_columns = set(source_columns.keys()) - set(existing_columns.keys())
missing_columns = set(existing_columns.keys()) - set(source_columns.keys())

Expected output to user:

Schema Comparison:

Existing table columns: customer_id, name, email
Source data columns: customer_id, name, email, phone

New columns in source (will be added): phone
Missing columns in source (will be NULL): None

Schema evolution will automatically add new columns to the table.
2. Add New Columns via ALTER TABLE

With AWS CLI:

aws athena start-query-execution \
  --query-string "ALTER TABLE \"catalog\".\"namespace\".\"table\" ADD COLUMNS (phone STRING)" \
  --query-execution-context Database=namespace \
  --result-configuration OutputLocation=s3://bucket/results/ \
  --region us-east-1
3. Handle Missing Columns

If source is missing columns that exist in the target table, two approaches:

Option 1: Use NULL for missing columns (recommended) — New rows will have NULL in these columns. Existing rows keep their values.

Option 2: Fail the import — Ensures data completeness. Requires source to have all columns.

Nested JSON Handling

Flatten vs Preserve Decision

When source data has nested structures:

{
  "order_id": 12345,
  "customer": {
    "customer_id": 789,
    "name": "John Doe",
    "email": "[email protected]"
  },
  "items": [
    {"product_id": 456, "quantity": 2, "price": 29.99}
  ]
}
Flattening Implementation

PySpark - Flatten Struct:

from pyspark.sql.functions import col

flattened_df = source_df.select(
    col("order_id"),
    col("customer.customer_id").alias("customer_id"),
    col("customer.name").alias("customer_name"),
    col("customer.email").alias("customer_email"),
    col("order_date"),
    col("total")
)

PySpark - Explode Array:

from pyspark.sql.functions import explode, col

# One row per item
exploded_df = source_df.select(
    col("order_id"),
    col("customer.customer_id").alias("customer_id"),
    explode(col("items")).alias("item")
).select(
    "order_id",
    "customer_id",
    col("item.product_id"),
    col("item.quantity"),
    col("item.price")
)

Athena SQL - Flatten with UNNEST:

-- Create external table with nested types
CREATE EXTERNAL TABLE orders_nested (
  order_id BIGINT,
  customer STRUCT<customer_id: BIGINT, name: STRING, email: STRING>,
  items ARRAY<STRUCT<product_id: BIGINT, quantity: INT, price: DECIMAL(10,2)>>,
  order_date DATE,
  total DECIMAL(10,2)
)
ROW FORMAT SERDE 'org.openx.data.jsonserde.JsonSerDe'
LOCATION 's3://bucket/orders/';

-- Flatten and insert
INSERT INTO "catalog"."namespace"."orders_flat"
SELECT
  order_id,
  customer.customer_id,
  customer.name AS customer_name,
  customer.email AS customer_email,
  item.product_id,
  item.quantity,
  item.price,
  order_date
FROM orders_nested
CROSS JOIN UNNEST(items) AS t(item);
Preserving Nested Structures

S3 Tables DDL with Nested Types:

CREATE TABLE "catalog"."namespace"."orders_nested" (
  order_id BIGINT,
  customer STRUCT<
    customer_id: BIGINT,
    name: STRING,
    email: STRING
  >,
  items ARRAY<STRUCT<
    product_id: BIGINT,
    quantity: INT,
    price: DECIMAL(10,2)
  >>,
  order_date DATE,
  total DECIMAL(10,2)
)
USING ICEBERG

Querying Nested Data:

-- Access struct fields
SELECT
  order_id,
  customer.name,
  customer.email,
  order_date
FROM "catalog"."namespace"."orders_nested"
WHERE customer.customer_id = 789;

-- Explode array in queries
SELECT
  order_id,
  item.product_id,
  item.quantity,
  item.price
FROM "catalog"."namespace"."orders_nested"
CROSS JOIN UNNEST(items) AS t(item);

PySpark - Write with Nested Types:

# Preserve nested structure
source_df.writeTo(args['target_table']).append()

# No flattening needed - PySpark DataFrame schema maps directly to Iceberg

Array Handling Options

Implementation examples for each array handling approach:

Option 1: Keep as Array

Store as ARRAY<STRUCT<...>> in S3 Table. Query with UNNEST when needed. Preserves one-to-many relationships efficiently.

Option 2: Explode to Separate Rows

Each array element becomes its own row. Simple flat table structure. May create many duplicate rows if arrays are large.

Option 3: Create Separate Related Table

Store items in separate table (e.g., order_items). Link via foreign key. Normalized database design.

Complete Examples

Example 1: Schema Evolution

Before (existing table):

CREATE TABLE customers (
  customer_id INT,
  name STRING,
  email STRING
)

New Source Data adds columns: phone STRING, address STRING

After Evolution:

ALTER TABLE customers ADD COLUMNS (
  phone STRING,
  address STRING
);

Result:

  • Existing rows: customer_id=1, name="Alice", email="[email protected]", phone=NULL, address=NULL

  • New rows: customer_id=2, name="Bob", email="[email protected]", phone="555-1234", address="123 Main St"

Example 2: Nested JSON with Flattening

Source JSON:

{
  "user_id": 100,
  "profile": {
    "age": 30,
    "city": "Seattle"
  },
  "purchases": [
    {"item": "book", "amount": 20},
    {"item": "laptop", "amount": 1200}
  ]
}

Flattened Table:

user_id | age | city    | item   | amount
--------|-----|---------|--------|-------
100     | 30  | Seattle | book   | 20
100     | 30  | Seattle | laptop | 1200

PySpark Code:

from pyspark.sql.functions import explode, col

df = spark.read.json("s3://bucket/data.json")

flattened = df.select(
    col("user_id"),
    col("profile.age"),
    col("profile.city"),
    explode(col("purchases")).alias("purchase")
).select(
    "user_id",
    "age",
    "city",
    col("purchase.item"),
    col("purchase.amount")
)
Example 3: Nested JSON Preserved

Same Source, but preserved as nested:

Table Schema:

CREATE TABLE user_purchases (
  user_id BIGINT,
  profile STRUCT<age: INT, city: STRING>,
  purchases ARRAY<STRUCT<item: STRING, amount: DECIMAL(10,2)>>
)

Query Example:

-- Get users from Seattle who bought laptops
SELECT
  user_id,
  profile.age,
  purchase.item,
  purchase.amount
FROM user_purchases
CROSS JOIN UNNEST(purchases) AS t(purchase)
WHERE profile.city = 'Seattle'
  AND purchase.item = 'laptop';

Evaluation Criteria

Schema Evolution

Detection:

  • Compares source schema to existing table schema

  • Identifies new, missing, and changed columns

  • Reports differences clearly to user

Automatic Handling:

  • New columns: Automatically executes ALTER TABLE ADD COLUMNS

  • Missing columns: Uses NULL or asks user

  • Type changes: Routes to type conflict resolution

Execution:

  • ALTER TABLE commands are syntactically correct

  • Uses appropriate Iceberg/S3 Tables syntax

  • Verifies changes applied successfully

Nested JSON

Detection:

  • Identifies STRUCT and ARRAY types in source

  • Determines nesting depth

  • Lists all nested fields clearly

User Choice:

  • Presents flatten vs preserve options

  • Explains pros/cons of each approach

  • Waits for user decision

Implementation:

  • Flatten: Provides complete PySpark/SQL with explode for arrays

  • Preserve: Creates correct DDL with nested types

  • Validates nested schema is correct

Query Examples:

  • Shows how to query nested data

  • Demonstrates struct field access (e.g., customer.name)

  • Shows UNNEST/explode for arrays

Common Mistakes to Avoid

Recreating entire table when only ALTER TABLE ADD COLUMNS is needed Silently using NULL for missing columns without informing user Not asking user how to handle nested structures (flatten vs preserve) Incomplete flattening code (missing some nested fields) Incorrect DDL for nested types (wrong syntax) Not validating that ALTER TABLE succeeded Exploding arrays without explaining it creates multiple rows Not providing query examples for nested data access

Requirements and compatibility

Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta

“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.

Needs: Needs an MCP server, Needs a command-line tool

Security

What our scanner noticed· Low risk

  • Instructions that use the network

A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.

Attribution and licence

Published by
AWS
Publisher
AWS (Official)
Source
aws/agent-toolkit-for-aws · plugins/aws-data-analytics/skills/ingesting-into-data-lake · commit 0a218f8
Licence
Apache License 2.0 (redistribution allowed)
  • LICENSE (added by AutoGineer from LICENSE)
  • NOTICE (added by AutoGineer from NOTICE)
  • ATTRIBUTION.md (written by AutoGineer)
Synced
Also distributed in
aws/agent-toolkit-for-aws

AutoGineer is not affiliated with or endorsed by AWS.

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download ingesting-into-data-lake.zip