Advanced Caching: Bloom Filter for Known-Bad Address Lookup

[0xVida]

Description:
Use a Bloom filter for constant-time lookups of known-bad Stellar addresses in the blocklist.

Requirements and Context:
Checking a DB table for every incoming address lookup under high TPS is too slow. A Bloom filter reduces this to O(1) with near-zero memory.

Acceptance Criteria:

• Bloom filter built from the blocklist on server startup
• Updated incrementally when new addresses are added to the blocklist
• False positive rate < 0.1% (configured via filter size)
• Benchmarks comparing filter vs DB lookup latency

Deliverables:

• Implementation of the above criteria
• Tests or proof of correct behaviour

NOTE: This issue will NOT be reviewed or approved without evidence (screenshot, log, or test output) demonstrating the feature works as described.

Suggested Execution:

1. Use the bloom-filters npm package
2. Pre-populate from the DB blocklist on startup
3. Subscribe to blocklist update events to rebuild the filter

Guidelines:

• Follow the existing project coding conventions
• All new environment variables must be documented in .env.example
• Phase: Phase 14: Performance

Example Commit Message:

perf: bloom filter for constant-time blocklist address lookup