admin/Nexus
1
0
Fork 0
Code Issues Pull Requests Actions 82 Packages Projects Releases Wiki Activity

多Worker守护 + 告警服务器名 + 清理PHP残留

Browse Source 
- main.py: Redis主Worker选举，防止多Worker重复执行后台任务
- agent.py: 告警/恢复广播使用真实服务器名(查MySQL)替代"server-{id}"
- sync_v2.py: 修复browse目录解析死代码，正确取parts[8]
- install.html: 移除过时install.php引用，更新为"删除.env重装"
- nginx配置: 移除PHP-FPM路由(install.html纯静态无需PHP)
- schedule_runner: 清理未使用的get_redis_sync导入

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
Your Name
2026-05-22 11:54:11 +08:00
parent b36e1d8010
commit cb5b4c42de
8 changed files with 77 additions and 47 deletions
Download Patch File Download Diff File Expand all files Collapse all files
deploy/nginx_http.conf
-15
View File
@@ -4,7 +4,6 @@
# {{SERVER_NAME}} = your domain or IP (e.g. nexus.example.com or 192.168.1.100)
# {{DEPLOY_DIR}} = installation root (e.g. /opt/nexus or /www/wwwroot/nexus.example.com)
# {{LOG_DIR}} = nginx log directory (e.g. /var/log/nginx or /www/wwwlogs)
# {{PHP_SOCK}} = PHP-FPM socket path (e.g. unix:/tmp/php-cgi-82.sock) — only if install.php is used
# ── Upstream: Python FastAPI backend ──
upstream nexus_api {
@@ -51,20 +50,6 @@ server {
proxy_set_header Host $host;
}
# ── install.php (PHP-FPM for installer, root = /web/) ──
location ~ ^/install\.php$ {
root {{DEPLOY_DIR}}/web;
fastcgi_pass {{PHP_SOCK}};
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
include fastcgi_params;
}
# ── Installer + Agent assets (root = /web/) ──
location ~ ^/(css|js|vendor|agent)/ {
root {{DEPLOY_DIR}}/web;
try_files $uri =404;
}
# ── Static assets ──
location ~* \.(css|js|png|jpg|jpeg|gif|ico|svg|woff|woff2|ttf|eot)$ {
expires 7d;
deploy/nginx_https.conf
-15
View File
@@ -6,7 +6,6 @@
# {{LOG_DIR}} = nginx log directory (e.g. /var/log/nginx or /www/wwwlogs)
# {{SSL_CERT}} = SSL certificate fullchain path
# {{SSL_KEY}} = SSL private key path
# {{PHP_SOCK}} = PHP-FPM socket path (e.g. unix:/tmp/php-cgi-82.sock) — only if install.php is used
# ── Upstream: Python FastAPI backend ──
upstream nexus_api {
@@ -42,20 +41,6 @@ server {
root {{DEPLOY_DIR}}/web/app;
index index.html;
# ── install.php (PHP-FPM for installer, root = /web/) ──
location ~ ^/install\.php$ {
root {{DEPLOY_DIR}}/web;
fastcgi_pass {{PHP_SOCK}};
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
include fastcgi_params;
}
# ── Installer + Agent assets (root = /web/) ──
location ~ ^/(css|js|vendor|agent)/ {
root {{DEPLOY_DIR}}/web;
try_files $uri =404;
}
# ── Python API proxy ──
location /api/ {
proxy_pass http://nexus_api;
server/api/agent.py
+11 -2
View File
@@ -115,6 +115,15 @@ async def receive_heartbeat(
logger.error(f"Redis heartbeat write failed for server {server_id}: {e}")
# Redis write failed — still continue with MySQL update
# ── 1b. Look up server name for alert broadcasts ──
server_name = f"server-{server_id}"
try:
server = await service.get_server(server_id)
if server:
server_name = server.name
except Exception:
pass
# ── 2. Alert detection ──
alert_thresholds = {
"cpu": settings.CPU_ALERT_THRESHOLD,
@@ -126,7 +135,7 @@ async def receive_heartbeat(
alerts = _detect_alerts(system_info, alert_thresholds)
for alert_type, alert_value in alerts:
logger.warning(f"Alert: server {server_id} {alert_type}={alert_value}%")
await broadcast_alert(server_id, alert_type, alert_value, f"server-{server_id}")
await broadcast_alert(server_id, alert_type, alert_value, server_name)
# Track active alert in Redis
try:
@@ -145,7 +154,7 @@ async def receive_heartbeat(
recoveries = _detect_recovery(system_info, prev_alerts, alert_thresholds)
for metric, value in recoveries:
logger.info(f"Recovery: server {server_id} {metric}={value}%")
await broadcast_recovery(server_id, metric, value, f"server-{server_id}")
await broadcast_recovery(server_id, metric, value, server_name)
# Remove recovered alert from Redis tracking
for prev in prev_alerts:
server/api/install.py
+1 -1
View File
@@ -27,7 +27,7 @@ router = APIRouter(prefix="/api/install", tags=["install"])
ROOT_DIR = Path(__file__).resolve().parent.parent.parent
ENV_FILE = ROOT_DIR / ".env"
LOCK_FILE = ROOT_DIR / "web" / "install.php.locked"
LOCK_FILE = ROOT_DIR / "web" / "install.php.locked" # Legacy compat — primary lock is .install_locked
INSTALL_LOCK = ROOT_DIR / ".install_locked"
CONFIG_PHP_DIR = ROOT_DIR / "web" / "data"
CONFIG_PHP = CONFIG_PHP_DIR / "config.php"
server/api/sync_v2.py
+1 -5
View File
@@ -135,13 +135,9 @@ async def browse_directory(
continue
parts = line.split(None, 8)
if len(parts) >= 9:
perms, _, owner, group, size, *date_parts, name = (
parts[0], parts[1], parts[2], parts[3], parts[4],
parts[5:-1], parts[-1]
)
is_dir = parts[0].startswith("d")
entries.append({
"name": parts[-1],
"name": parts[8], # Last field after split(None, 8) — handles spaces in names
"is_dir": is_dir,
"size": parts[4],
"perms": parts[0],
server/background/schedule_runner.py
-1
View File
@@ -8,7 +8,6 @@ import logging
from datetime import datetime, timezone
from server.infrastructure.database.session import AsyncSessionLocal
from server.infrastructure.redis.client import get_redis_sync
from server.domain.models import PushSchedule, SyncLog, AuditLog
logger = logging.getLogger("nexus.schedule_runner")
server/main.py
+62 -6
View File
@@ -67,6 +67,55 @@ logger = logging.getLogger("nexus")
# Track background tasks for clean shutdown
_background_tasks: list[asyncio.Task] = []
# Redis-based primary worker lock (prevents duplicate background tasks with multi-worker uvicorn)
_PRIMARY_LOCK_KEY = "nexus:primary_worker"
_PRIMARY_LOCK_TTL = 30 # seconds — must renew periodically
async def _acquire_primary_lock() -> bool:
"""Try to become the primary worker via Redis SETNX.
Only one uvicorn worker will hold this lock at a time.
The lock auto-expires after PRIMARY_LOCK_TTL seconds (safety net for crashes).
Primary worker renews the lock periodically via _renew_primary_lock().
"""
try:
from server.infrastructure.redis.client import get_redis
import os
redis = get_redis()
worker_id = str(os.getpid())
# SET with NX (only if not exists) + EX (TTL)
acquired = await redis.set(_PRIMARY_LOCK_KEY, worker_id, nx=True, ex=_PRIMARY_LOCK_TTL)
if acquired:
# Start lock renewal background task
asyncio.create_task(_renew_primary_lock(), name="primary_lock_renewal")
logger.info(f"Primary worker lock acquired (pid={worker_id})")
return True
return False
except Exception as e:
# If Redis fails, assume primary (single-worker mode)
logger.warning(f"Primary lock check failed, assuming primary: {e}")
return True
async def _renew_primary_lock():
"""Periodically renew the primary worker lock in Redis."""
import os
worker_id = str(os.getpid())
while True:
await asyncio.sleep(_PRIMARY_LOCK_TTL // 2) # Renew at half the TTL
try:
from server.infrastructure.redis.client import get_redis
redis = get_redis()
current = await redis.get(_PRIMARY_LOCK_KEY)
if current == worker_id:
await redis.expire(_PRIMARY_LOCK_KEY, _PRIMARY_LOCK_TTL)
else:
logger.warning("Primary worker lock lost — another worker took over")
break
except Exception as e:
logger.error(f"Primary lock renewal failed: {e}")
# ── D7: DB Session Middleware (fixes session leak in 4 service factories) ──
@@ -150,12 +199,19 @@ async def lifespan(app: FastAPI):
from server.api.websocket import start_redis_subscriber, stop_redis_subscriber
await start_redis_subscriber()
# 5. Launch background tasks
task_flush = asyncio.create_task(heartbeat_flush_loop(), name="heartbeat_flush")
task_monitor = asyncio.create_task(self_monitor_loop(), name="self_monitor")
task_schedule = asyncio.create_task(schedule_runner_loop(), name="schedule_runner")
task_retry = asyncio.create_task(retry_runner_loop(), name="retry_runner")
_background_tasks.extend([task_flush, task_monitor, task_schedule, task_retry])
# 5. Launch background tasks (only on primary worker to avoid duplicate execution)
# When running with --workers N, each worker gets its own lifespan.
# Use Redis-based leader election to ensure only one worker runs background tasks.
is_primary = await _acquire_primary_lock()
if is_primary:
task_flush = asyncio.create_task(heartbeat_flush_loop(), name="heartbeat_flush")
task_monitor = asyncio.create_task(self_monitor_loop(), name="self_monitor")
task_schedule = asyncio.create_task(schedule_runner_loop(), name="schedule_runner")
task_retry = asyncio.create_task(retry_runner_loop(), name="retry_runner")
_background_tasks.extend([task_flush, task_monitor, task_schedule, task_retry])
logger.info("Primary worker — background tasks launched")
else:
logger.info("Secondary worker — background tasks skipped (primary worker handles them)")
# 6. Start asyncssh connection pool (W1)
from server.infrastructure.ssh.asyncssh_pool import ssh_pool
web/app/install.html
+2 -2
View File
@@ -314,7 +314,7 @@
<div class="w-16 h-16 rounded-full bg-green-100 text-green-500 text-3xl flex items-center justify-center mx-auto mb-4">✓</div>
<h2 class="text-xl font-bold text-slate-800 mb-2">安装完成！</h2>
<p class="text-slate-500 mb-1">Nexus 6.0 已成功安装。</p>
<p class="text-slate-400 text-xs">install.php 已重命名为 install.php.locked — 防止重复安装</p>
<p class="text-slate-400 text-xs">安装向导已锁定 — 防止重复安装（如需重装请删除 .env 文件后重启服务）</p>
<!-- Guardian results -->
<div x-show="initResult?.guardian_results?.length" class="text-left mt-6 mb-4 rounded-lg p-4"
@@ -457,7 +457,7 @@ location / {
安全收尾
</div>
<div class="text-xs text-slate-500 space-y-1">
<p>✓ <code class="bg-slate-800 text-slate-200 px-1.5 py-0.5 rounded">install.php</code> 已自动重命名为 <code class="bg-slate-800 text-slate-200 px-1.5 py-0.5 rounded">install.php.locked</code></p>
<p>✓ 安装向导已锁定 — 删除 <code class="bg-slate-800 text-slate-200 px-1.5 py-0.5 rounded">.env</code> 文件可重新安装</p>
<p>✓ <code class="bg-slate-800 text-slate-200 px-1.5 py-0.5 rounded">.env</code> 包含 SECRET_KEY/API_KEY — <b>安装后不可修改</b>（加密一致性）</p>
<p>✓ 防火墙限制端口 <span x-text="form.api_port"></span> (仅允许本机和子服务器IP)</p>
<p>✓ 修改管理员默认密码</p>