Affected: jahlives openssl_encrypt

openssl_encrypt versions before 1.4.0 use an in-memory rate limiter for TOTP brute-force protection that is not shared across workers and is lost on server restart. Attackers can distribute authentication attempts across multiple server instances or retry immediately after a restart to bypass rate limiting protections.

Affected: PyPI: openssl-encrypt

## Severity: HIGH

### Summary

The TOTP brute-force rate limiter in `openssl_encrypt_server/modules/pepper/totp.py` at **lines 47-98** uses an in-memory `defaultdict(list)` as a class variable.

### Affected Code

```python class TOTPRateLimiter: def __init__(self, ...): self.attempts: Dict[str, List[datetime]] = defaultdict(list) self.lockouts: Dict[str, datetime] = {}

class TOTPService: _rate_limiter = TOTPRateLimiter() # Class variable, in-memory only ```

### Impact

1. Rate limit state is **not shared** across multiple server instances/workers — an attacker can distribute attempts 2. All rate limit state is **lost on server restart** — allows immediate retry 3. In multi-worker deployments, each worker has independent rate limit state

### Recommended Fix

- Use Redis or the database for rate limit state storage - Or use a shared-memory approach for multi-worker deployments - At minimum, persist lockout state to survive restarts

### Fix

Fixed in commit `2749bc0` on branch `releases/1.4.x` — added abstract RateLimitBackend with InMemoryBackend and DatabaseBackend implementations; defaults to DatabaseBackend when DB available.