Update performance report with summary of all fixes applied

Updated executive summary to document:
- 29 files modified total
- 6 categories of fixes applied
- Specific models and improvements listed
This commit is contained in:
Claude 2026-01-07 19:22:19 +00:00
parent 600e5b71fb
commit 051d10fe34
No known key found for this signature in database

View File

@ -2,14 +2,27 @@
## Executive Summary
This report identifies **72+ instances** of performance anti-patterns across the dbt project. The most critical issues involve:
This report identifies **72+ instances** of performance anti-patterns across the dbt project and documents the fixes applied.
### Issues Identified:
1. **Missing incremental predicates** on 72 of 86 incremental models (84%)
2. **N+1 query patterns** with 58 `run_query()` calls across 32 files
3. **Non-sargable JOIN conditions** using `LOWER()` function calls
4. **Cartesian product** in observability model
5. **Repeated subquery execution** for `MAX(_inserted_timestamp)`
### Fixes Applied (29 files modified):
1. ✅ **silver__transactions.sql** - Reduced `run_query()` calls from 4 to 1 (75% reduction)
2. ✅ **silver__nft_sales_combined.sql** - Cached `MAX(_inserted_timestamp)` as Jinja variable
3. ✅ **silver_observability__transactions_completeness.sql** - Fixed Cartesian product with explicit `CROSS JOIN`
4. ✅ **silver__hourly_prices_priority.sql** - Pre-computed `LOWER()` in CTEs for sargable JOINs
5. ✅ **core__ez_transfers.sql** - Pre-computed `LOWER()` and `DATE_TRUNC()` in CTEs
6. ✅ **24 silver models** - Added `incremental_predicates` for partition pruning:
- 17 DEX swap models
- 3 NFT sales models
- 2 NFT mint models
- 2 bridge transfer models
---
## Critical Issues