Skip to content

Latest commit

 

History

History
67 lines (55 loc) · 4 KB

File metadata and controls

67 lines (55 loc) · 4 KB
title Developing Code for Data Processing
type category
tags
data-engineer-professional
developing-code
data-processing
status published

Developing Code for Data Processing (22 % of Exam)

The highest-weighted domain in the November 30, 2025 blueprint. Covers writing and optimising Python and SQL for batch and streaming workloads, Delta Lake operations, Lakeflow Declarative Pipelines, Lakeflow Jobs, and the workspace/notebook tooling that hosts the code.

Topics Overview

flowchart TD
    DC[Developing Code] --> Batch[Batch ETL]
    DC --> Inc[Incremental Processing]
    DC --> Stream[Structured Streaming]
    DC --> Delta[Delta Lake Operations]
    DC --> SJ[Streaming Joins & Stateful Ops]
    DC --> LDP[Lakeflow Declarative Pipelines]
    DC --> LJ[Lakeflow Jobs]
    DC --> WS[Workspace & Notebooks]
Loading

Section Contents

File Topic Priority
01-batch-etl-pipelines-part1.md ETL design patterns, DataFrame transformations, joins, aggregations High
01-batch-etl-pipelines-part2.md Performance optimisation, error handling, use cases, exam tips High
02-incremental-processing.md Incremental loads, checkpoint management High
03-structured-streaming-part1.md Streaming fundamentals, triggers, output modes, watermarking High
03-structured-streaming-part2.md Stream-static joins, stateful operations, state store, exam tips High
04-delta-lake-operations-part1.md MERGE, OPTIMIZE, VACUUM, ZORDER, time travel, table cloning High
04-delta-lake-operations-part2.md Schema operations, table properties, Delta 3.0+ features, exam tips High
05-streaming-joins-stateful.md Stream-stream joins, stateful ops, watermarking, deduplication High
06-declarative-pipelines.md Lakeflow Declarative Pipelines syntax, materialized views, streaming tables High
07-lakeflow-jobs-part1.md Lakeflow Jobs basics, tasks, dependencies, parameters High
07-lakeflow-jobs-part2.md Job triggers, advanced patterns, monitoring, exam tips High
08-workspace-and-notebooks.md Workspace organisation, notebook patterns, magic commands Medium

Key Concepts to Master

Concept Why it matters
Idempotent writes Re-running the same code must produce the same output (critical for retries)
Checkpointing Streaming queries persist state to durable storage for fault tolerance
Watermarking Bounds how long state is kept for late-arriving data in streaming joins/aggregations
Delta MERGE Single-statement upsert that handles inserts, updates, and deletes atomically
Lakeflow Declarative Pipelines Declarative ETL framework (formerly Lakeflow Declarative Pipelines) — define the what, not the how
Lakeflow Jobs Orchestration layer (formerly Lakeflow Jobs) for scheduling notebooks, pipelines, and Python wheels

Related Resources


↑ Back to DE Professional | Next: Cost & Performance Optimization → (first domain — no previous)