Optimize Apache Iceberg table design and maintenance through Trino.
npx skills add https://github.com/ivanshamaev/de-agent-skills --skill trino-iceberg
Please help me install this Agent Skill. Skill: trino_iceberg Source: https://github.com/ivanshamaev/de-agent-skills/tree/main/skills/trino_iceberg Command: npx skills add https://github.com/ivanshamaev/de-agent-skills --skill trino-iceberg
This Skill helps you create, query efficiently, and safely maintain Apache Iceberg tables when working through Trino, preventing slow scans caused by poor partitioning, missing stats, or incorrect maintenance workflows.
Use the trino_iceberg skill to diagnose a slow Iceberg query by running EXPLAIN (TYPE DISTRIBUTED) and then deciding whether to change partitioning, add bucketing/sorted_by, and apply ANALYZE or table maintenance.
High-intent search queries and answers about installing and using this skill.
To optimize slow Trino queries on Apache Iceberg tables, run EXPLAIN (TYPE DISTRIBUTED) to diagnose scan bottlenecks, then adjust partition transforms, add bucketing or sorted_by properties, and apply ANALYZE to update table statistics for improved partition pruning.
Run maintenance procedures like compaction and snapshot expiration on Iceberg tables in Trino by executing OPTIMIZE for data compaction, EXPIRE_SNAPSHOTS to remove old snapshots, and REMOVE_ORPHAN_FILES to clean up unreferenced data files.
Time travel query works for rollback in Trino Iceberg by reading historical table snapshots, allowing you to query past data states for recovery and audit purposes without altering the current table data.
Yes, you can perform schema evolution and partition updates on Iceberg tables through Trino using ALTER TABLE statements, which support adding columns, evolving schemas, and updating table properties to adapt to changing data requirements.
Iceberg partition pruning scans in Trino remain slow when table statistics are missing or outdated, requiring you to run ANALYZE and verify metadata table diagnostics like $partitions and $snapshots to ensure correct partition transform behavior.
The best way to diagnose Iceberg table metadata issues in Trino is to query metadata tables like $partitions and $snapshots, which provide detailed insights into partition distribution and snapshot history for troubleshooting table maintenance and performance.
ivanshamaevDesign and maintain high-performance Apache Iceberg tables in Trino.
weirdgmeDesign and implement scalable platform engineering patterns for internal developer platforms.
IndustrialDesign and implement batch and streaming data pipelines with ETL, dbt, Spark, Airflow, and Delta Lake.
snoodleboot-ioDiagnose slow SQL queries with EXPLAIN ANALYZE and apply indexing optimizations.
