ClaudeSuperPower

aidp-spark-optimization

Skill

Use when a Spark job/notebook is slow, missing an SLA, spilling, OOMing, generating too many small files, or shuffling/skewing heavily; when reviewing Spark/PySpark code or a Spark UI for performance; or before running a large Spark workload. Covers open-source Apache Spark 3.5.0 (+ Delta Lake) tuni

Install

git clone https://github.com/oracle-samples/oracle-aidp-samples.git ~/.claude/skills/aidp-spark-optimization

What is aidp-spark-optimization?

Use when a Spark job/notebook is slow, missing an SLA, spilling, OOMing, generating too many small files, or shuffling/skewing heavily; when reviewing Spark/PySpark code or a Spark UI for performance; or before running a large Spark workload. Covers open-source Apache Spark 3.5.0 (+ Delta Lake) tuning -- partitions, shuffle, joins, skew, file layout, memory, codegen, caching, AQE, compression, and configuration.

What this can do

Capabilities declared in this component's own frontmatter — not inferred.

Inherit all session tools

Declares no tool restrictions — inherits every session tool

~104 tokens of context used while enabled, before you invoke anything

Documentation

README · ~6 min read

Spark Optimization

Field-tested techniques for optimizing open-source Apache Spark 3.5.0 (and open-source Delta Lake) workloads. Every technique lists what it does, why it matters, patterns (when it helps), anti-patterns (when it does not help or hurts), the exact config/code, and impact evidence from real reproductions.

Applies to plain OSS Spark 3.5.0. Vendor-only features (Databricks Photon / predictive optimization / Delta disk cache, AWS Glue/EMR runtime tricks) are excluded or explicitly flagged. (Note: Delta Lake liquid clustering, OPTIMIZE/ZORDER, deletion vectors, CDF are open-source — GA in Delta 3.2.0 — see references/08-delta-lake.md.) Some examples run on the Oracle AI Data Platform (AIDP); AIDP-specific execution rules are in references/aidp-notes.md.

In this plugin (oracle-ai-data-platform-workbench-engineer-agent): measure/apply on AIDP via the bundled scripts/aidp_sql.py (one SparkSession per cluster -- set/revert spark.conf explicitly; see references/aidp-notes.md); pull Spark-UI stage/task metrics via the aidp-spark-debugging skill; right-size compute + set cluster-create-only configs (e.g. spark.memory.fraction) via aidp-cluster-ops; run Delta OPTIMIZE/VACUUM/ZORDER DDL via aidp-sql-ddl / aidp-table-management. Adapted from the ai-data-engineer-agent spark-optimization skill (Oracle).

Core principle

Reviews

Log in to leave a review.

No reviews yet — be the first.

Explore related

Other things in this space — across every part of the ecosystem, not just skills.