降低iDRAC会话压力并开放重试配置

.env.example

@@ -2,3 +2,9 @@ HOST=192.168.1.100
 USERNAME=root
 PASSWORD=your_idrac_password
 FAN_SPEED_STEPS=50:20,55:25,60:30,65:40
+CONTROL_INTERVAL_SECONDS=120
+ERROR_INTERVAL_SECONDS=120
+IPMI_RETRY_COUNT=5
+IPMI_RETRY_DELAY_SECONDS=20
+IPMI_TIMEOUT_SECONDS=60
+USE_RAW_FAN_DUTY=false

CHANGES.md

@@ -10,6 +10,7 @@
 - 清理 Python 缓存文件，避免将运行产物提交到仓库。
 - Docker 运行镜像切换到 Debian slim，只安装 `python3`、`ipmitool` 和时区数据，在降低体积的同时保留更好的IPMI兼容性。
 - 新增 `FAN_SPEED_STEPS` 环境变量，允许用户通过 `.env` 自定义温度阈值和风扇转速档位。
+- 新增轮询间隔、IPMI重试、命令超时和 raw 风扇占空比查询开关，默认减少 iDRAC8 的会话压力。
 
 ## Previous Improvements
 

README.md

@@ -69,6 +69,12 @@ HOST=192.168.1.100
 USERNAME=root
 PASSWORD=your_idrac_password
 FAN_SPEED_STEPS=50:20,55:25,60:30,65:40
+CONTROL_INTERVAL_SECONDS=120
+ERROR_INTERVAL_SECONDS=120
+IPMI_RETRY_COUNT=5
+IPMI_RETRY_DELAY_SECONDS=20
+IPMI_TIMEOUT_SECONDS=60
+USE_RAW_FAN_DUTY=false
 ```
 
 Start the service / 启动服务：
@@ -108,11 +114,21 @@ python3 start.py
 | `USERNAME` | Yes | iDRAC username with IPMI permission. / 有 IPMI 权限的 iDRAC 用户名。 |
 | `PASSWORD` | Yes | iDRAC password. / iDRAC 密码。 |
 | `FAN_SPEED_STEPS` | No | Temperature-to-speed rules. Default: `50:20,55:25,60:30,65:40`. / 温度和风扇转速规则，默认值：`50:20,55:25,60:30,65:40`。 |
+| `CONTROL_INTERVAL_SECONDS` | No | Normal control interval. Default: `120`. / 正常控制间隔，默认 `120` 秒。 |
+| `ERROR_INTERVAL_SECONDS` | No | Wait time after a failed control cycle. Default: same as `CONTROL_INTERVAL_SECONDS`. / 控制周期失败后的等待时间，默认等于正常控制间隔。 |
+| `IPMI_RETRY_COUNT` | No | Retry count for each IPMI command. Default: `5`. / 单条 IPMI 命令重试次数，默认 `5`。 |
+| `IPMI_RETRY_DELAY_SECONDS` | No | Wait time between IPMI retries. Default: `20`. / IPMI 重试间隔，默认 `20` 秒。 |
+| `IPMI_TIMEOUT_SECONDS` | No | Subprocess timeout for each IPMI command. Default: `60`. / 单次 IPMI 命令超时时间，默认 `60` 秒。 |
+| `USE_RAW_FAN_DUTY` | No | Query raw fan duty before RPM estimation. Default: `false`. / 是否先用 raw 命令读取风扇占空比，默认 `false`。 |
 
 The application does not include default credentials. Missing variables will stop startup with a clear error.
 
 程序不内置默认地址、账号或密码。缺少环境变量时会直接停止并输出明确错误。
 
+For iDRAC8 systems with unstable IPMI sessions, keep `CONTROL_INTERVAL_SECONDS` at `120` or higher. The default `USE_RAW_FAN_DUTY=false` avoids one extra IPMI session per cycle and estimates fan percentage from the RPM data already returned by `sdr`.
+
+对于 IPMI 会话不稳定的 iDRAC8，建议 `CONTROL_INTERVAL_SECONDS` 保持 `120` 秒或更高。默认 `USE_RAW_FAN_DUTY=false` 会跳过额外的 raw 占空比查询，直接使用同一次 `sdr` 返回的 RPM 估算风扇百分比，减少 iDRAC 会话压力。
+
 ## Temperature Policy / 温控策略
 
 The controller reads all temperature sensors and uses the highest value.

controller/client.py

@@ -15,13 +15,31 @@ DEFAULT_FAN_SPEED_STEPS = (
 class FanController:
 
     # 初始化控制器并记录iDRAC连接信息
-    def __init__(self, host: str, username: str, password: str, fan_speed_steps: str = None):
+    def __init__(
+        self,
+        host: str,
+        username: str,
+        password: str,
+        fan_speed_steps: str = None,
+        ipmi_retry_count: int = 5,
+        ipmi_retry_delay: int = 20,
+        ipmi_timeout: int = 60,
+        use_raw_fan_duty: bool = False,
+    ):
         self.host = host
         self.username = username
         self.password = password
 
-        self.ipmi = IpmiTool(self.host, self.username, self.password)
+        self.ipmi = IpmiTool(
+            self.host,
+            self.username,
+            self.password,
+            retry_count=ipmi_retry_count,
+            retry_delay=ipmi_retry_delay,
+            timeout=ipmi_timeout,
+        )
         self.fan_speed_steps = self.parse_fan_speed_steps(fan_speed_steps)
+        self.use_raw_fan_duty = use_raw_fan_duty
         self.last_set_speed = None  # 记录最后设置的风扇速度
         self.is_auto_mode = False   # 记录当前是否为自动模式
 
@@ -116,7 +134,7 @@ class FanController:
                 self.is_auto_mode = False
 
             # 获取当前风扇转速
-            current_speed = self.ipmi.get_fan_duty_cycle(sensor_data)
+            current_speed = self.ipmi.get_fan_duty_cycle(sensor_data, use_raw=self.use_raw_fan_duty)
 
             # 只有在当前转速与所需转速不同时才调整
             # 如果无法获取当前转速（返回-1），则检查是否已记录之前设置的速度

controller/ipmi.py

@@ -7,31 +7,33 @@ from controller.logger import logger
 # IPMI命令封装器：负责调用ipmitool读取传感器并设置Dell风扇
 class IpmiTool:
     # 初始化iDRAC连接参数
-    def __init__(self, host: str, username: str, password: str):
+    def __init__(self, host: str, username: str, password: str, retry_count: int = 5, retry_delay: int = 20, timeout: int = 60):
         if not host or not username or not password:
             raise ValueError("host, username and password must be provided")
         self.host = host
         self.username = username
         self.password = password
+        self.retry_count = retry_count
+        self.retry_delay = retry_delay
+        self.timeout = timeout
 
     # 执行ipmitool命令并处理重试、超时和会话异常
     def run_cmd(self, cmd: str) -> str:
         basecmd = f'ipmitool -H {self.host} -I lanplus -U {self.username} -P {self.password}'
         command = f'{basecmd} {cmd}'
-        retry_count = 5  # 增加重试次数以应对网络波动
+        for attempt in range(self.retry_count):
-        for attempt in range(retry_count):
             try:
                 # print(f"Executing command: {command}")  # 添加调试信息
-                result = subprocess.run(command, shell=True, capture_output=True, text=True, timeout=60)  # 增加超时时间
+                result = subprocess.run(command, shell=True, capture_output=True, text=True, timeout=self.timeout)  # 控制单次命令最长等待时间
 
                 if result.returncode != 0:
                     # 部分ipmitool版本会把错误输出到stdout，或只返回退出码但stderr为空
                     error_msg = result.stderr.strip() or result.stdout.strip() or f'命令退出码: {result.returncode}'
                     # 检查是否是网络连接问题
                     if "Unable to establish IPMI" in error_msg or "session" in error_msg:
-                        logger.warning(f'IPMI会话建立失败 (尝试 {attempt + 1}/{retry_count}): {error_msg}')
+                        logger.warning(f'IPMI会话建立失败 (尝试 {attempt + 1}/{self.retry_count}): {error_msg}')
-                        if attempt < retry_count - 1:
+                        if attempt < self.retry_count - 1:
-                            time.sleep(10)  # 网络问题需要更长的等待时间
+                            time.sleep(self.retry_delay)  # 网络问题需要更长的等待时间
                             continue
                     raise RuntimeError(
                         f'IPMI 命令执行失败: {cmd}\n错误详情: {error_msg}'  # 更清晰的错误提示
@@ -39,16 +41,16 @@ class IpmiTool:
 
                 return result.stdout
             except subprocess.TimeoutExpired:
-                logger.warning(f'命令超时 (尝试 {attempt + 1}/{retry_count})')
+                logger.warning(f'命令超时 (尝试 {attempt + 1}/{self.retry_count})')
-                if attempt < retry_count - 1:
+                if attempt < self.retry_count - 1:
-                    logger.warning(f'正在重试... (尝试次数 {attempt + 1}/{retry_count})')
+                    logger.warning(f'正在重试... (尝试次数 {attempt + 1}/{self.retry_count})')
-                    time.sleep(10)  # 每次重试前等待更长时间
+                    time.sleep(self.retry_delay)  # 每次重试前等待更长时间
                 else:
                     raise RuntimeError('IPMI 命令超时。请检查网络连接或服务器状态。')  # 更明确的错误提示
             except Exception as e:
-                logger.warning(f'IPMI命令执行异常 (尝试 {attempt + 1}/{retry_count}): {str(e)}')
+                logger.warning(f'IPMI命令执行异常 (尝试 {attempt + 1}/{self.retry_count}): {str(e)}')
-                if attempt < retry_count - 1:
+                if attempt < self.retry_count - 1:
-                    time.sleep(10)  # 网络问题需要更长的等待时间
+                    time.sleep(self.retry_delay)  # 网络问题需要更长的等待时间
                 else:
                     raise e
 
@@ -116,29 +118,30 @@ class IpmiTool:
         return fan_speeds
 
     # 获取当前风扇占空比，raw命令不可用时用RPM估算
-    def get_fan_duty_cycle(self, sensor_data: str = None) -> int:
+    def get_fan_duty_cycle(self, sensor_data: str = None, use_raw: bool = False) -> int:
         """
         获取当前风扇占空比/百分比
         :return: current fan duty cycle in percentage
         """
-        try:
+        if use_raw:
-            # Raw command to get current fan duty cycle
+            try:
-            result = self.run_cmd('raw 0x30 0x31 0x01')
+                # Raw command to get current fan duty cycle
-            # Parse the hex result to get duty cycle
+                result = self.run_cmd('raw 0x30 0x31 0x01')
-            result_parts = result.strip().split()
+                # Parse the hex result to get duty cycle
-            if result_parts and len(result_parts) >= 1:
+                result_parts = result.strip().split()
-                # The command should return a hex value representing the duty cycle
+                if result_parts and len(result_parts) >= 1:
-                duty_cycle_hex = result_parts[-1]
+                    # The command should return a hex value representing the duty cycle
-                duty_cycle = int(duty_cycle_hex, 16)
+                    duty_cycle_hex = result_parts[-1]
-                # Ensure the value is in valid range (0-100)
+                    duty_cycle = int(duty_cycle_hex, 16)
-                if 0 <= duty_cycle <= 100 and duty_cycle != 0:
+                    # Ensure the value is in valid range (0-100)
-                    # If we get a reasonable value (not 0), return it
+                    if 0 <= duty_cycle <= 100 and duty_cycle != 0:
-                    return duty_cycle
+                        # If we get a reasonable value (not 0), return it
-                elif duty_cycle == 0:
+                        return duty_cycle
-                    # Value of 0 might indicate auto mode or that raw command doesn't return duty cycle on this system
+                    elif duty_cycle == 0:
-                    logger.info('原始命令返回0，尝试从RPM估算风扇百分比')
+                        # Value of 0 might indicate auto mode or that raw command doesn't return duty cycle on this system
-        except Exception as e:
+                        logger.info('原始命令返回0，尝试从RPM估算风扇百分比')
-            logger.warning(f'获取风扇占空比的原始命令失败: {e}')
+            except Exception as e:
+                logger.warning(f'获取风扇占空比的原始命令失败: {e}')
 
         # If raw command fails or returns 0, get fan speeds from sensor data and convert to approximate percentage
         try:

start.py

@@ -5,6 +5,40 @@ import traceback
 from controller.client import FanController
 from controller.logger import logger
 
+
+# 解析整数环境变量，缺省时使用默认值
+def get_int_env(name: str, default: int, min_value: int = 1) -> int:
+    value = os.getenv(name)
+    if value is None:
+        return default
+
+    try:
+        parsed_value = int(value)
+    except ValueError as exc:
+        raise RuntimeError(f'{name} 必须是整数') from exc
+
+    if parsed_value < min_value:
+        raise RuntimeError(f'{name} 必须大于等于 {min_value}')
+
+    return parsed_value
+
+
+# 解析布尔环境变量，支持 true/false、1/0、yes/no
+def get_bool_env(name: str, default: bool = False) -> bool:
+    value = os.getenv(name)
+    if value is None:
+        return default
+
+    normalized_value = value.strip().lower()
+    if normalized_value in ('1', 'true', 'yes', 'on'):
+        return True
+
+    if normalized_value in ('0', 'false', 'no', 'off'):
+        return False
+
+    raise RuntimeError(f'{name} 必须是布尔值: true/false')
+
+
 if __name__ == '__main__':
 
     # 从环境变量读取iDRAC连接信息，开源版本不内置任何真实默认凭据
@@ -15,6 +49,12 @@ if __name__ == '__main__':
     fan_speed_steps = os.getenv('FAN_SPEED_STEPS')
     if fan_speed_steps is None:
         fan_speed_steps = os.getenv('FAN_SPEED_RULES')
+    control_interval = get_int_env('CONTROL_INTERVAL_SECONDS', 120)
+    error_interval = get_int_env('ERROR_INTERVAL_SECONDS', control_interval)
+    ipmi_retry_count = get_int_env('IPMI_RETRY_COUNT', 5)
+    ipmi_retry_delay = get_int_env('IPMI_RETRY_DELAY_SECONDS', 20)
+    ipmi_timeout = get_int_env('IPMI_TIMEOUT_SECONDS', 60)
+    use_raw_fan_duty = get_bool_env('USE_RAW_FAN_DUTY', False)
     if not host:
         raise RuntimeError('未设置 HOST 环境变量')
 
@@ -30,16 +70,20 @@ if __name__ == '__main__':
         username=username,
         password=password,
         fan_speed_steps=fan_speed_steps,
+        ipmi_retry_count=ipmi_retry_count,
+        ipmi_retry_delay=ipmi_retry_delay,
+        ipmi_timeout=ipmi_timeout,
+        use_raw_fan_duty=use_raw_fan_duty,
     )
 
     while True:
         try:            
             # 执行一次温度读取和风扇控制周期
             client.run()
-            time.sleep(60)
+            time.sleep(control_interval)
         except Exception as err:
             logger.error(
                 f'运行控制器失败 {err}. {traceback.format_exc()}'
             )
             # iDRAC会话异常时等待下一轮，避免连续请求压垮IPMI服务
-            time.sleep(60)
+            time.sleep(error_interval)

tests/test_fan_controller.py

@@ -49,6 +49,11 @@ class FanControllerConfigTest(unittest.TestCase):
                 with self.assertRaises(ValueError):
                     self.make_controller(invalid_value)
 
+    def test_raw_fan_duty_query_is_disabled_by_default(self):
+        controller = self.make_controller()
+
+        self.assertFalse(controller.use_raw_fan_duty)
+
 
 if __name__ == '__main__':
     unittest.main()

tests/test_start_config.py

@@ -0,0 +1,32 @@
+import os
+import unittest
+from unittest.mock import patch
+
+from start import get_bool_env, get_int_env
+
+
+class StartConfigTest(unittest.TestCase):
+    def test_get_int_env_uses_default_when_missing(self):
+        with patch.dict(os.environ, {}, clear=True):
+            self.assertEqual(get_int_env('CONTROL_INTERVAL_SECONDS', 120), 120)
+
+    def test_get_int_env_rejects_invalid_value(self):
+        with patch.dict(os.environ, {'CONTROL_INTERVAL_SECONDS': 'abc'}, clear=True):
+            with self.assertRaises(RuntimeError):
+                get_int_env('CONTROL_INTERVAL_SECONDS', 120)
+
+    def test_get_bool_env_parses_supported_values(self):
+        with patch.dict(os.environ, {'USE_RAW_FAN_DUTY': 'true'}, clear=True):
+            self.assertTrue(get_bool_env('USE_RAW_FAN_DUTY'))
+
+        with patch.dict(os.environ, {'USE_RAW_FAN_DUTY': 'false'}, clear=True):
+            self.assertFalse(get_bool_env('USE_RAW_FAN_DUTY'))
+
+    def test_get_bool_env_rejects_invalid_value(self):
+        with patch.dict(os.environ, {'USE_RAW_FAN_DUTY': 'maybe'}, clear=True):
+            with self.assertRaises(RuntimeError):
+                get_bool_env('USE_RAW_FAN_DUTY')
+
+
+if __name__ == '__main__':
+    unittest.main()