[program:webapi]
command=/opt/venv/bin/python /app/main.py
directory=/app
user=appuser
autostart=true
autorestart=true
startsecs=3
startretries=5

stdout_logfile=/var/log/webapi.out.log
stdout_logfile_maxbytes=50MB
stdout_logfile_backups=10

stderr_logfile=/var/log/webapi.err.log
stderr_logfile_maxbytes=50MB
stderr_logfile_backups=10

environment=PYTHONPATH="/app",PRODUCTION="1"

apiVersion: apps/v1
kind: Deployment
metadata:
  name: data-processor
spec:
  replicas: 3
  selector:
    matchLabels:
      app: data-processor
  template:
    metadata:
      labels:
        app: data-processor
    spec:
      containers:
      - name: main
        image: registry.example.com/data-processor:v1.2.3
        resources:
          limits:
            cpu: "2"
            memory: 4Gi
          requests:
            cpu: "1"
            memory: 2Gi
        livenessProbe:
          exec:
            command: ["python", "/app/healthcheck.py"]
          initialDelaySeconds: 30
          periodSeconds: 10
        readinessProbe:
          httpGet:
            path: /health
            port: 8080
        volumeMounts:
          - name: config-volume
            mountPath: /app/config
      volumes:
        - name: config-volume
          configMap:
            name: app-config

# 分布式鎖示例（Redis實現(xiàn)）
import redis
from redis.lock import Lock

class HAWorker:
    def __init__(self):
        self.redis = redis.Redis(host='redis-cluster', port=6379)
        self.lock_name = "task:processor:lock"
        
    def run(self):
        while True:
            with Lock(self.redis, self.lock_name, timeout=30, blocking_timeout=5):
                self.process_data()
                
            time.sleep(1)
            
    def process_data(self):
        # 核心業(yè)務(wù)邏輯
        pass

# 基于Prometheus的存活檢測
from prometheus_client import start_http_server, Gauge

class HeartbeatMonitor:
    def __init__(self, port=9000):
        self.heartbeat = Gauge('app_heartbeat', 'Last successful heartbeat')
        start_http_server(port)
        
    def update(self):
        self.heartbeat.set_to_current_time()

# 在業(yè)務(wù)代碼中集成
monitor = HeartbeatMonitor()
while True:
    process_data()
    monitor.update()
    time.sleep(60)

# 使用psutil進行資源監(jiān)控
import psutil

def monitor_resources():
    return {
        "cpu_percent": psutil.cpu_percent(interval=1),
        "memory_used": psutil.virtual_memory().used / 1024**3,
        "disk_io": psutil.disk_io_counters().read_bytes,
        "network_io": psutil.net_io_counters().bytes_sent
    }

# 集成到Prometheus exporter
from prometheus_client import Gauge

cpu_gauge = Gauge('app_cpu_usage', 'CPU usage percentage')
mem_gauge = Gauge('app_memory_usage', 'Memory usage in GB')

def update_metrics():
    metrics = monitor_resources()
    cpu_gauge.set(metrics['cpu_percent'])
    mem_gauge.set(metrics['memory_used'])

# 創(chuàng)建專用用戶
sudo useradd -r -s /bin/false appuser

# 設(shè)置文件權(quán)限
sudo chown -R appuser:appgroup /opt/app
sudo chmod 750 /opt/app

# 使用capabilities替代root
sudo setcap CAP_NET_BIND_SERVICE=+eip /opt/venv/bin/python

# 使用seccomp限制系統(tǒng)調(diào)用
import prctl

def enable_sandbox():
    # 禁止fork新進程
    prctl.set_child_subreaper(1)
    prctl.set_no_new_privs(1)
    
    # 限制危險系統(tǒng)調(diào)用
    from seccomp import SyscallFilter, ALLOW, KILL
    filter = SyscallFilter(defaction=KILL)
    filter.add_rule(ALLOW, "read")
    filter.add_rule(ALLOW, "write")
    filter.add_rule(ALLOW, "poll")
    filter.load()

# 基于檢查點的狀態(tài)恢復
import pickle
from datetime import datetime

class StateManager:
    def __init__(self):
        self.state_file = "/var/run/app_state.pkl"
        
    def save_state(self, data):
        with open(self.state_file, 'wb') as f:
            pickle.dump({
                'timestamp': datetime.now(),
                'data': data
            }, f)
            
    def load_state(self):
        try:
            with open(self.state_file, 'rb') as f:
                return pickle.load(f)
        except FileNotFoundError:
            return None

# 在業(yè)務(wù)邏輯中集成
state_mgr = StateManager()
last_state = state_mgr.load_state()

while True:
    process_data(last_state)
    state_mgr.save_state(current_state)
    time.sleep(60)

# AWS多區(qū)域部署示例
resource "aws_instance" "app_east" {
  provider = aws.us-east-1
  ami           = "ami-0c55b159cbfafe1f0"
  instance_type = "t3.large"
  count         = 3
}

resource "aws_instance" "app_west" {
  provider = aws.us-west-2
  ami           = "ami-0c55b159cbfafe1f0"
  instance_type = "t3.large"
  count         = 2
}

resource "aws_route53_record" "app" {
  zone_id = var.dns_zone
  name    = "app.example.com"
  type    = "CNAME"
  ttl     = "300"
  records = [
    aws_lb.app_east.dns_name,
    aws_lb.app_west.dns_name
  ]
}

# 使用__slots__減少內(nèi)存占用
class DataPoint:
    __slots__ = ['timestamp', 'value', 'quality']
    
    def __init__(self, ts, val, q):
        self.timestamp = ts
        self.value = val
        self.quality = q

# 使用memory_profiler分析
@profile
def process_data():
    data = [DataPoint(i, i*0.5, 1) for i in range(1000000)]
    return sum(d.value for d in data)

# 使用Cython加速
# File: fastmath.pyx
cimport cython

@cython.boundscheck(False)
@cython.wraparound(False)
def calculate(double[:] array):
    cdef double total = 0.0
    cdef int i
    for i in range(array.shape[0]):
        total += array[i] ** 2
    return total

# 使用multiprocessing并行
from multiprocessing import Pool

def parallel_process(data_chunks):
    with Pool(processes=8) as pool:
        results = pool.map(process_chunk, data_chunks)
    return sum(results)

# AWS Lambda函數(shù)示例
import boto3

def lambda_handler(event, context):
    s3 = boto3.client('s3')
    
    # 處理S3事件
    for record in event['Records']:
        bucket = record['s3']['bucket']['name']
        key = record['s3']['object']['key']
        
        # 執(zhí)行處理邏輯
        process_file(bucket, key)
        
    return {
        'statusCode': 200,
        'body': 'Processing completed'
    }

# 基于機器學習異常檢測
from sklearn.ensemble import IsolationForest

class AnomalyDetector:
    def __init__(self):
        self.model = IsolationForest(contamination=0.01)
        
    def train(self, metrics_data):
        self.model.fit(metrics_data)
        
    def predict(self, current_metrics):
        return self.model.predict([current_metrics])[0]

# 集成到監(jiān)控系統(tǒng)
detector = AnomalyDetector()
detector.train(historical_metrics)

current = collect_metrics()
if detector.predict(current) == -1:
    trigger_alert()