Cloudera Data Operator study plan

Map the Cloudera Data Platform architecture: list the control plane and data plane components and write a one-sentence role for each.

Compare CDP Public Cloud and CDP Private Cloud Base on at least six dimensions (deployment, storage, compute, governance, upgrade model, cost model) in a table.

Trace a single workload request from client to storage in CDP and annotate each hop with the service responsible.

Explain how Cloudera Manager, Cloudera Data Warehouse, and Cloudera Data Engineering divide administrative responsibility.

Configure and validate a small CDP environment end to end in a sandbox: create the environment, attach storage, and launch one service.

Differentiate the shared-responsibility boundaries between Cloudera and the customer for security, patching, and backup.

Diagnose three common environment-provisioning failures (credential, network, storage permission) and document the fix for each.

Summarize the CDP release cadence and how runtime versions are selected per cluster.

Describe the Cloudera AI (formerly CML) workspace model: projects, runtimes, sessions, jobs, and models, with the lifecycle of each.

Build and run a small model-training job in a workspace and register the resulting model in the model registry.

Explain how model governance works: lineage, model versions, approval states, and auditability.

Compare batch scoring via jobs with real-time model deployment via inference endpoints and state when each is appropriate.

Connect an AI workspace to governed data through the shared data fabric so that access policies are enforced at query time.

Demonstrate row-level or column-level masking on a dataset used by a model-training job and verify the model sees only permitted data.

Explain how metadata, lineage, and classification flow from the data fabric into AI experiments.

Identify the governance controls required before an AI model can consume production data and map each to a platform feature.

Describe the Data in Motion stack: ingestion, streaming, schema registry, and stream processing, and the role of each layer.

Build a flow that ingests events, validates them against a schema, and lands them in a downstream table.

Explain delivery guarantees (at-least-once vs. exactly-once), checkpointing, and how consumer offsets are managed.

Compare NiFi-based flow management with Kafka-based streaming for three concrete ingestion scenarios.

Explain the open data lakehouse layout: table formats, catalog, file formats, and the compute engines that read them.

Create a managed table, perform an update and delete, and show how the table format handles snapshots and time travel.

Contrast open table formats on schema evolution, partitioning, and concurrency, and state which fits a given workload.

Query the same lakehouse table from two different engines and confirm consistent results and shared metadata.

Map the unified data fabric: catalog, lineage, policy, and access control, and how they interoperate across services.

Apply a classification tag to a dataset and verify that the tag drives an access or masking policy automatically.

Trace end-to-end lineage from an ingested event through a lakehouse table to a report or model, and document each node.

Explain how the fabric enforces consistent policy across batch, streaming, and AI workloads without per-service duplication.

Assemble a full-stack mini-project that ingests events, stores them in a lakehouse table, trains a model on governed data, and serves predictions.

Audit the project against all five domains and list one gap per domain with a concrete remediation step.

Complete two timed practice sets weighted 20 percent per domain and review every missed item to root cause.

Write a one-page exam-day reference covering the highest-risk facts you repeatedly missed.