Update overview.rst by tdcmeehan · Pull Request #1 · tdcmeehan/presto

tdcmeehan · 2023-08-04T17:26:39Z

Testing

Edited query.yaml and created cluster.yaml edited the cluster.yml created queuestatement.yaml Add detailed OpenAPI schema for BasicQueryInfo and other updates Update queued_statement.yaml Updated query.yaml and added _full_spec.yaml Split and add missing components (#1) Also: - re-adds full_spec.yaml to gitignore, - Adds execution to generate a single openAPI JSON. The output is a single JSON that can be used with the swagger editor.

…ies and complex subqueries Issue #1 (cyclic join graphs): RPT's LargestRoot algorithm identifies back edges that create cycles. Materializing those joins as CTEs decomposes the cyclic graph into acyclic sections, each of which RPT can fully reduce independently. Issue #3 (complex subquery join inputs): Materializing aggregation/nested-join subqueries as CTEs converts them to TableScanNodes on temporary tables. RPT can then push BFs into these scans with column pruning. CteProjectionAndPredicatePushDown can further push TupleDomain ranges back into CTE producers. Both leverage existing CTE infrastructure: LogicalCteOptimizer, PhysicalCteOptimizer, SequenceNode, CTEMaterializationTracker. https://claude.ai/code/session_01SAXk4AS8yQyQkhRVi6RQie

New issues: - #13: Cost model assumes local NVMe, invalid for S3/HDFS (0.3s becomes 5-30s) - prestodb#14: Parquet BF section is 115-line dead end, should be condensed - prestodb#15: Duplicate section numbering (two "Section 6" headers) - prestodb#16: Superlinear join efficiency claims lack citations - prestodb#17: CTE materialization cost missing from quantitative cost model - prestodb#18: Reduced hash table may still exceed VRAM — fallback composition unclear Updated existing issues: #1,#3 remaining gaps noted, #2 now worse (more occurrences), #7 partially addressed. 12 open issues total. https://claude.ai/code/session_01SAXk4AS8yQyQkhRVi6RQie

Update overview.rst

65adec2