Error Handling in Trading Systems: Why It Matters More Than You Think

# DANGEROUS -- never do this in trading code
try:
    place_order(symbol, direction, size)
except:
    pass  # "It's fine, we'll catch it next time"
    # Narrator: it was not fine

import logging
import requests

logger = logging.getLogger("trading")

def fetch_indicator_safe(symbol: str, indicator: str, timeframe: str) -> dict:
    """Fetch indicator with explicit error handling."""
    try:
        resp = requests.get(
            "https://tickatlas.com/v1/indicator",
            headers={"X-API-Key": API_KEY},
            params={"symbol": symbol, "indicator": indicator, "timeframe": timeframe},
            timeout=10,
        )
        resp.raise_for_status()
        data = resp.json()

        if not data.get("success"):
            error = data.get("error", {})
            logger.error(f"API error for {symbol}/{indicator}: {error}")
            raise ValueError(f"API returned error: {error}")

        return data["data"]

    except requests.exceptions.Timeout:
        logger.error(f"Timeout fetching {indicator} for {symbol}")
        raise  # Let caller decide what to do

    except requests.exceptions.ConnectionError:
        logger.error(f"Connection failed fetching {indicator} for {symbol}")
        raise

    except requests.exceptions.HTTPError as e:
        logger.error(f"HTTP error {e.response.status_code} for {symbol}/{indicator}")
        if e.response.status_code == 429:
            raise  # Rate limit -- caller should back off
        raise

import time

class CircuitBreaker:
    def __init__(self, failure_threshold: int = 5, recovery_timeout: int = 60):
        self.failure_threshold = failure_threshold
        self.recovery_timeout = recovery_timeout
        self.failure_count = 0
        self.last_failure_time = 0
        self.state = "CLOSED"  # CLOSED = normal, OPEN = failing

    def can_execute(self) -> bool:
        if self.state == "CLOSED":
            return True
        # Check if recovery timeout has passed
        if time.time() - self.last_failure_time > self.recovery_timeout:
            self.state = "HALF_OPEN"
            return True
        return False

    def record_success(self):
        self.failure_count = 0
        self.state = "CLOSED"

    def record_failure(self):
        self.failure_count += 1
        self.last_failure_time = time.time()
        if self.failure_count >= self.failure_threshold:
            self.state = "OPEN"
            logger.warning("Circuit breaker OPEN -- pausing API calls")

# Usage
breaker = CircuitBreaker(failure_threshold=3, recovery_timeout=30)

def safe_fetch(symbol: str, indicator: str, timeframe: str):
    if not breaker.can_execute():
        logger.warning("Circuit breaker open -- using last known value")
        return get_cached_value(symbol, indicator, timeframe)

    try:
        data = fetch_indicator_safe(symbol, indicator, timeframe)
        breaker.record_success()
        return data
    except Exception:
        breaker.record_failure()
        raise

def get_signal_with_fallback(symbol: str) -> dict:
    """Try primary signal, fall back to simpler analysis."""
    # Try full multi-indicator analysis
    try:
        rsi = fetch_indicator_safe(symbol, "RSI_14", "H4")
        macd = fetch_indicator_safe(symbol, "MACD_hist", "H4")
        return analyze_confluence(rsi, macd)
    except Exception as e:
        logger.warning(f"Full analysis failed for {symbol}: {e}")

    # Fall back to single indicator
    try:
        rsi = fetch_indicator_safe(symbol, "RSI_14", "H4")
        # The API returns the number; the buy/sell call is yours to make.
        value = rsi["value"]
        signal = "buy" if value < 30 else "sell" if value > 70 else "hold"
        return {"signal": signal, "confidence": "low", "degraded": True}
    except Exception as e:
        logger.error(f"All analysis failed for {symbol}: {e}")

    # Final fallback: no signal
    return {"signal": "no_data", "confidence": "none", "degraded": True}

import redis

r = redis.Redis()

def heartbeat():
    """Call this every loop iteration."""
    r.setex("bot:heartbeat", 120, "alive")  # Expires in 2 minutes

# Separate monitoring script (cron every 2 minutes):
def check_heartbeat():
    if not r.exists("bot:heartbeat"):
        send_alert("CRITICAL: Trading bot has stopped responding!")
        close_all_positions()  # Emergency flatten