IBM watsonx Data Lakehouse Engineer Associate study plan
IBM watsonx Data Lakehouse Engineer Associate · Give each week a clear focus.
8 weeks · About 6 hours per week
watsonx.data architecture and provisioning
Define watsonx.data's lakehouse architecture: separate compute and storage, the metastore, and the roles of Presto, Spark, and Db2.
Identify the provisioning options (Lite, Standard, Enterprise) and list which services and limits each tier includes.
Diagram the network and identity flow from a client through the query engine to object storage.
Provision a Lite instance and record the catalog, engine, and bucket defaults it creates.
Your checkpoint
A one-page architecture diagram with annotations plus a written comparison table of the three provisioning tiers.
watsonx.data architecture and provisioning & Data ingestion and connection management
Provision a Standard instance and configure an object storage bucket as a new catalog.
Register Hive and Iceberg catalogs and verify table visibility from the query workspace.
Create a schema and table, then confirm the underlying files appear in the bucket.
Compare native versus external catalog registration and record when each is appropriate.
Your checkpoint
A working catalog in your instance plus a short runbook documenting every configuration value you set.
Data ingestion and connection management
Connect watsonx.data to at least two source types (for example Db2 and a cloud object store) and ingest sample tables.
Configure ingestion with the ingestion job UI and schedule a recurring load.
Write a data-quality check that rejects malformed rows and log the outcome.
Explain how schema evolution is handled for Iceberg tables during ingestion.
Your checkpoint
Two ingested tables with a scheduled job and a log showing a successful and a rejected load.
Data ingestion and connection management & Query engines and performance optimization
Run the same query on Presto and Spark and compare runtime and resource consumption.
Create a partitioned Iceberg table and measure query time against an unpartitioned copy.
Tune a slow query by rewriting joins and filtering before aggregation.
Document when to choose Presto versus Spark for a given workload.
Your checkpoint
A benchmark table of at least four queries with before-and-after timings and a one-paragraph recommendation.
Query engines and performance optimization
Configure an engine's resource groups or worker sizing and observe the effect on concurrent queries.
Use EXPLAIN to identify a full table scan and fix it with partitioning or file compaction.
Compact small files in an Iceberg table and verify the reduction in file count.
Summarize the main query-performance levers available in watsonx.data.
Your checkpoint
A performance-tuning report covering one engine configuration change and one storage optimization, each with measured impact.
Security, governance, and monitoring
Map platform roles to least-privilege access and assign them to two test users.
Configure storage and catalog permissions so one user can read but not write a table.
Enable audit logging and locate the entries for a query and a configuration change.
Describe how watsonx.data integrates with an external identity provider for authentication.
Your checkpoint
A role-permission matrix and a redacted audit log excerpt showing a denied write attempt.
Security, governance, and monitoring & watsonx.data architecture and provisioning
Set up monitoring for engine health and query failures and define two alert conditions.
Apply data governance metadata (tags or descriptions) to a table and confirm it is visible to consumers.
Review cost drivers such as idle engines and storage growth and propose two controls.
Revisit the architecture diagram and annotate where security and monitoring controls sit.
Your checkpoint
An operations checklist covering monitoring, alerting, governance tagging, and cost controls, with evidence of each being exercised.
watsonx.data architecture and provisioning & Data ingestion and connection management & Query engines and performance optimization & Security, governance, and monitoring
Complete a timed 60-question practice set spanning all four domains and score at least 80%.
Review every missed question and write the underlying fact in your own words.
Re-provision a clean instance and repeat the core setup from memory in under 45 minutes.
Produce a one-page cheat sheet of configuration defaults, role names, and tuning rules.
Your checkpoint
A scored practice exam with a written remediation list and a final one-page cheat sheet.
Milestones
Week 2: You can provision an instance and register a working catalog without consulting notes.
Week 4: You have ingested data from two source types and can justify Presto versus Spark for a given query.
Week 6: You can demonstrate least-privilege access with a denied write and show the matching audit entry.
Week 8: You score at least 80% on a timed mixed-domain practice set and can rebuild the core environment from memory.