feat(database): optimize timescaledb queries with continuous aggregat…

[DevNetlife]

Description
Over 90 days of operation, the prices, health_scores, and liquidity_snapshots hypertables accumulate millions of rows. Aggregating data over long ranges ($\ge 7$ days, 30 days, 90 days) on the Analytics dashboard previously executed expensive raw table queries.

This PR introduces TimescaleDB continuous aggregate materialized views with automated refresh policies for 1-hour and 1-day windows across high-volume hypertables, and refactors backend services to route queries to continuous aggregates when requesting data ranges $\ge 7$ days.

Type of Change
Optimization (performance enhancement)
Database Migration
Refactoring (no logic breaking changes)
Key Changes

1. Database Migration
   Created migration 048_continuous_aggregates_optimizations.ts defining TimescaleDB continuous aggregate materialized views:
   prices_hourly & prices_daily
   health_scores_hourly & health_scores_daily
   liquidity_hourly & liquidity_daily
   Configured automated refresh policies via add_continuous_aggregate_policy for 1-hour and 1-day bucket intervals.
   Added performance indexes on (symbol, bucket DESC) for each aggregate view.
2. Service Refactoring
   PriceModel (backend/src/database/models/price.model.ts): Refactored getTimeBucketed() to target prices_daily for lookbacks $\ge 30$ days and prices_hourly for lookbacks $\ge 7$ days, falling back to raw hypertables if unaggregated or in test environments.
   AnalyticsService (backend/src/services/analytics.service.ts): Refactored getHistoricalComparison() for health_score, liquidity, and price metrics to query continuous aggregates when requested range is $\ge 7$ days.
   AggregationService (backend/src/services/aggregation.service.ts): Refactored aggregatePrices(), aggregateHealthScores(), and aggregateVolume() to query continuous aggregate views for long time ranges ($\ge 7$ days) or daily/weekly/monthly intervals.
   LiquidityFragmentationService (backend/src/services/liquidityFragmentation.service.ts): Refactored getFragmentationTrend() for 7d and 30d ranges to query liquidity_hourly and liquidity_daily.
3. Testing & Verification
   Added test file backend/tests/services/continuousAggregates.service.test.ts to verify query routing logic for $\ge 7$-day ranges.
   Confirmed all service unit tests pass cleanly:
   Test Files 5 passed (5)
   Tests 79 passed (79)
   Impact & Compatibility
   Breaking Change: No
   Short-term Feeds: Unaffected (queries under 7 days continue to read directly from raw hypertables).

Closes #798

[Mosas2000]

Fix the failing E2E check

[Mosas2000]

The implementation of TimescaleDB continuous aggregates is excellently structured, featuring robust fallback mechanisms that will significantly enhance long-range query performance.

[Mosas2000]

The thorough test coverage effectively validates the newly optimized routing logic.

Thank you, approving this for merge.