Is there an existing issue for this?
Describe the bug
I used the zero-downtime restart solution. The new process was launched, while the old process became a defunct process. Tell me how to solve it? thank you
sanic 23.12
Python 3.11.10
Code snippet
import os
import signal
import sys
from sanic.worker.constants import RestartOrder
async def setup_restart_listener(app, loop):
# 只有主进程且非 Windows 系统才注册 SIGHUP
# hasattr(app, "m") 确保是在主管理进程中运行
if sys.platform == "win32":
return print("Windows 模式启动：信号监听已跳过。")
def handle_restart(sig, frame):
try:
# 1. 尝试使用 app.m
if hasattr(app, "m") and app.m:
print(f"主进程 {os.getpid()} 收到信号，通过 Multiplexer 触发...")
app.m.restart(zero_downtime=True)
# 2. 检查是否存在 manager 实例 (这是最稳健的访问方式)
elif hasattr(app, "manager") and app.manager:
print(f"主进程 {os.getpid()} 收到信号，通过 Manager 触发零停机重启...")
app.manager.restart(zero_downtime=True, restart_order=RestartOrder.STARTUP_FIRST)
else:
print("无法触发重启：Sanic Manager 尚未初始化完成")
except Exception as e:
print(f"触发重启失败: {str(e)}")
try:
signal.signal(signal.SIGHUP, handle_restart)
print(f"主进程 {os.getpid()} 已监听 SIGHUP (Zero-Downtime Mode)")
except Exception as e:
print(f"信号注册失败: {e}")
Expected Behavior
No response
How do you run Sanic?
As a module
Operating System
Linux
Sanic Version
23.12
Additional context
No response
Is there an existing issue for this?
Describe the bug
I used the zero-downtime restart solution. The new process was launched, while the old process became a defunct process. Tell me how to solve it? thank you
sanic 23.12
Python 3.11.10
Code snippet
import os
import signal
import sys
from sanic.worker.constants import RestartOrder
async def setup_restart_listener(app, loop):
# 只有主进程且非 Windows 系统才注册 SIGHUP
# hasattr(app, "m") 确保是在主管理进程中运行
Expected Behavior
No response
How do you run Sanic?
As a module
Operating System
Linux
Sanic Version
23.12
Additional context
No response