perf(db): add indexes on query-hot columns and Redis caching layer

- Add database indexes on FileRecord.created_at, FileRecord.mime_type,
  ProcessingLog.file_id, ProcessingLog.timestamp, and
  FileProcessingStep.status for faster filtering, sorting, and joins.
- Add _ensure_indexes() migration for existing databases.
- Create app/utils/cache.py with fail-open Redis GET/SET/DELETE helpers.
- Cache MIME types dropdown query in files view (120s TTL).
- Optimize batch status query to load only needed columns.

Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>
This commit is contained in:
copilot-swe-agent[bot]
2026-03-01 14:38:28 +00:00
parent a4fafe5c56
commit b79ba81e05
7 changed files with 467 additions and 10 deletions
+30
View File
@@ -165,6 +165,36 @@ def _run_schema_migrations(engine: Any) -> None:
conn.execute(text("CREATE INDEX IF NOT EXISTS ix_saved_searches_user_id ON saved_searches (user_id)"))
logger.info("Migration complete: 'saved_searches' table created")
# Migration: Add performance indexes for common query patterns
_ensure_indexes(engine, inspector)
def _ensure_indexes(engine: Any, inspector: Any) -> None:
"""
Create performance indexes for common query patterns.
Each ``CREATE INDEX IF NOT EXISTS`` is idempotent and safe to run
on every startup. The indexes target the columns most frequently
used in file listing/filtering, status computation and log retrieval.
"""
from sqlalchemy import text
_PERF_INDEXES = [
("ix_files_created_at", "files", "created_at"),
("ix_files_mime_type", "files", "mime_type"),
("ix_processing_logs_file_id", "processing_logs", "file_id"),
("ix_processing_logs_timestamp", "processing_logs", "timestamp"),
("ix_file_processing_steps_status", "file_processing_steps", "status"),
]
table_names = inspector.get_table_names()
with engine.begin() as conn:
for idx_name, table, column in _PERF_INDEXES:
if table in table_names:
conn.execute(text(f"CREATE INDEX IF NOT EXISTS {idx_name} ON {table} ({column})"))
logger.info("Performance indexes ensured")
def get_db() -> Generator[Session, None, None]:
"""