3 Commits
v1.3.6 ... main

Author SHA1 Message Date
xiaochao
0f691e956f 修复可用空间不可用问题 2025-07-28 14:37:28 +08:00
xiaochao
25348fff9b 优化硬盘检测逻辑避免唤醒休眠硬盘 2025-07-28 14:10:23 +08:00
xiaochao
30b1b7d271 修复ha中关闭飞牛系统电脑报错问题 2025-07-28 13:50:09 +08:00
4 changed files with 378 additions and 124 deletions

View File

@@ -313,11 +313,34 @@ class FlynasCoordinator(DataUpdateCoordinator):
if connection_id is not None:
await self.release_ssh_connection(connection_id)
async def ping_system(self) -> bool:
"""轻量级系统状态检测"""
# 对于本地主机直接返回True
if self.host in ['localhost', '127.0.0.1']:
return True
try:
# 使用异步ping检测减少超时时间
proc = await asyncio.create_subprocess_exec(
'ping', '-c', '1', '-W', '1', self.host,
stdout=asyncio.subprocess.DEVNULL,
stderr=asyncio.subprocess.DEVNULL
)
await asyncio.wait_for(proc.wait(), timeout=2) # 总超时时间2秒
return proc.returncode == 0
except Exception:
return False
async def _monitor_system_status(self):
"""系统离线时轮询检测状态"""
self._debug_log(f"启动系统状态监控,每{self._retry_interval}秒检测一次")
# 使用指数退避策略,避免频繁检测
check_interval = self._retry_interval
max_interval = 300 # 最大5分钟检测一次
while True:
await asyncio.sleep(self._retry_interval)
await asyncio.sleep(check_interval)
if await self.ping_system():
self._info_log("检测到系统已开机,触发重新加载")
@@ -326,24 +349,10 @@ class FlynasCoordinator(DataUpdateCoordinator):
self.hass.config_entries.async_reload(self.config_entry.entry_id)
)
break
async def ping_system(self) -> bool:
"""轻量级系统状态检测"""
# 对于本地主机直接返回True
if self.host in ['localhost', '127.0.0.1']:
return True
try:
# 使用异步ping检测
proc = await asyncio.create_subprocess_exec(
'ping', '-c', '1', '-W', '1', self.host,
stdout=asyncio.subprocess.DEVNULL,
stderr=asyncio.subprocess.DEVNULL
)
await proc.wait()
return proc.returncode == 0
except Exception:
return False
else:
# 系统仍然离线,增加检测间隔(指数退避)
check_interval = min(check_interval * 1.5, max_interval)
self._debug_log(f"系统仍离线,下次检测间隔: {check_interval}")
async def _async_update_data(self):
"""数据更新入口,优化命令执行频率"""
@@ -421,6 +430,14 @@ class FlynasCoordinator(DataUpdateCoordinator):
return self.get_default_data()
async def shutdown_system(self):
"""关闭系统 - 委托给SystemManager"""
return await self.system_manager.shutdown_system()
async def reboot_system(self):
"""重启系统 - 委托给SystemManager"""
return await self.system_manager.reboot_system()
class UPSDataUpdateCoordinator(DataUpdateCoordinator):
def __init__(self, hass: HomeAssistant, config, main_coordinator):
self.config = config

View File

@@ -14,6 +14,7 @@ class DiskManager:
self.disk_full_info_cache = {} # 缓存磁盘完整信息
self.first_run = True # 首次运行标志
self.initial_detection_done = False # 首次完整检测完成标志
self.disk_io_stats_cache = {} # 缓存磁盘I/O统计信息
def extract_value(self, text: str, patterns, default="未知", format_func=None):
if not text:
@@ -38,10 +39,9 @@ class DiskManager:
async def check_disk_active(self, device: str, window: int = 30) -> bool:
"""检查硬盘在指定时间窗口内是否有活动"""
try:
# 正确的路径是 /sys/block/{device}/stat
stat_path = f"/sys/block/{device}/stat"
# 读取统计文件
# 读取当前统计文件
stat_output = await self.coordinator.run_command(f"cat {stat_path} 2>/dev/null")
if not stat_output:
self.logger.debug(f"无法读取 {stat_path},默认返回活跃状态")
@@ -53,51 +53,147 @@ class DiskManager:
self.logger.debug(f"无效的统计信息格式:{stat_output}")
return True
# 关键字段当前正在进行的I/O操作数量第9个字段索引8
in_flight = int(stats[8])
try:
# /sys/block/{device}/stat 字段说明:
# 0: read I/Os requests 读请求次数
# 1: read I/Os merged 读请求合并次数
# 2: read sectors 读扇区数
# 3: read ticks 读操作耗时(ms)
# 4: write I/Os requests 写请求次数
# 5: write I/Os merged 写请求合并次数
# 6: write sectors 写扇区数
# 7: write ticks 写操作耗时(ms)
# 8: in_flight 当前进行中的I/O请求数
# 9: io_ticks I/O活动时间(ms)
# 10: time_in_queue 队列中的总时间(ms)
# 如果当前有I/O操作直接返回活跃状态
if in_flight > 0:
current_stats = {
'read_ios': int(stats[0]),
'write_ios': int(stats[4]),
'in_flight': int(stats[8]),
'io_ticks': int(stats[9])
}
# 如果当前有正在进行的I/O操作直接返回活跃状态
if current_stats['in_flight'] > 0:
self.logger.debug(f"磁盘 {device} 有正在进行的I/O操作: {current_stats['in_flight']}")
self.disk_io_stats_cache[device] = current_stats
return True
# 检查是否有缓存的统计信息
cached_stats = self.disk_io_stats_cache.get(device)
if cached_stats:
# 比较I/O请求次数的变化
read_ios_diff = current_stats['read_ios'] - cached_stats['read_ios']
write_ios_diff = current_stats['write_ios'] - cached_stats['write_ios']
io_ticks_diff = current_stats['io_ticks'] - cached_stats['io_ticks']
self.logger.debug(f"磁盘 {device} I/O变化: 读={read_ios_diff}, 写={write_ios_diff}, 活动时间={io_ticks_diff}ms")
# 如果在检测窗口内有I/O活动认为磁盘活跃
if read_ios_diff > 0 or write_ios_diff > 0 or io_ticks_diff > 100: # 100ms内的活动
self.logger.debug(f"磁盘 {device} 在窗口期内有I/O活动")
self.disk_io_stats_cache[device] = current_stats
return True
# 检查io_ticks是否表明最近有活动
# io_ticks是累积值如果在合理范围内增长说明有轻微活动
if io_ticks_diff > 0 and io_ticks_diff < window * 1000: # 在窗口时间内的轻微活动
self.logger.debug(f"磁盘 {device} 有轻微I/O活动")
self.disk_io_stats_cache[device] = current_stats
return True
else:
# 首次检测,保存当前状态并认为活跃
self.logger.debug(f"磁盘 {device} 首次检测,保存统计信息")
self.disk_io_stats_cache[device] = current_stats
return True
# 更新缓存
self.disk_io_stats_cache[device] = current_stats
# 检查硬盘电源状态
power_state = await self.get_disk_power_state(device)
if power_state in ["standby", "sleep", "idle"]:
self.logger.debug(f"磁盘 {device} 处于省电状态: {power_state}")
return False
# 所有检查都通过,返回非活跃状态
self.logger.debug(f"磁盘 {device} 判定为非活跃状态")
return False
except (ValueError, IndexError) as e:
self.logger.debug(f"解析统计信息失败: {e}")
return True
# 检查I/O操作时间第10个字段索引9 - io_ticks单位毫秒
io_ticks = int(stats[9])
# 如果设备在窗口时间内有I/O活动返回活跃状态
if io_ticks > window * 1000:
return True
# 所有检查都通过,返回非活跃状态
return False
except Exception as e:
self.logger.error(f"检测硬盘活动状态失败: {str(e)}", exc_info=True)
self.logger.error(f"检测硬盘活动状态失败: {str(e)}")
return True # 出错时默认执行检测
async def get_disk_activity(self, device: str) -> str:
"""获取硬盘活动状态(活动中/空闲中/休眠中)"""
async def get_disk_power_state(self, device: str) -> str:
"""获取硬盘电源状态"""
try:
# 检查硬盘是否处于休眠状态
# 检查 SCSI 设备状态
state_path = f"/sys/block/{device}/device/state"
state_output = await self.coordinator.run_command(f"cat {state_path} 2>/dev/null || echo 'unknown'")
state = state_output.strip().lower()
if state in ["standby", "sleep"]:
if state in ["running", "active"]:
return "active"
elif state in ["standby", "sleep"]:
return state
# 对于某些设备尝试通过hdparm检查状态非侵入性
hdparm_output = await self.coordinator.run_command(f"hdparm -C /dev/{device} 2>/dev/null || echo 'unknown'")
if "standby" in hdparm_output.lower():
return "standby"
elif "sleeping" in hdparm_output.lower():
return "sleep"
elif "active/idle" in hdparm_output.lower():
return "active"
return "unknown"
except Exception as e:
self.logger.debug(f"获取磁盘 {device} 电源状态失败: {e}")
return "unknown"
async def get_disk_activity(self, device: str) -> str:
"""获取硬盘活动状态(活动中/空闲中/休眠中)"""
try:
# 先检查电源状态
power_state = await self.get_disk_power_state(device)
if power_state in ["standby", "sleep"]:
return "休眠中"
# 检查最近一分钟内的硬盘活动
# 检查最近的I/O活动
stat_path = f"/sys/block/{device}/stat"
stat_output = await self.coordinator.run_command(f"cat {stat_path}")
stats = stat_output.split()
stat_output = await self.coordinator.run_command(f"cat {stat_path} 2>/dev/null")
if len(stats) >= 11:
# 第9个字段是最近完成的读操作数
# 第10个字段是最近完成的写操作数
recent_reads = int(stats[8])
recent_writes = int(stats[9])
if stat_output:
stats = stat_output.split()
if len(stats) >= 11:
try:
in_flight = int(stats[8]) # 当前进行中的I/O
if recent_reads > 0 or recent_writes > 0:
return "活动中"
# 如果有正在进行的I/O返回活动中
if in_flight > 0:
return "活动中"
# 检查缓存的统计信息来判断近期活动
cached_stats = self.disk_io_stats_cache.get(device)
if cached_stats:
current_read_ios = int(stats[0])
current_write_ios = int(stats[4])
read_diff = current_read_ios - cached_stats.get('read_ios', 0)
write_diff = current_write_ios - cached_stats.get('write_ios', 0)
if read_diff > 0 or write_diff > 0:
return "活动中"
except (ValueError, IndexError):
pass
return "空闲中"

View File

@@ -1,7 +1,7 @@
{
"domain": "fn_nas",
"name": "飞牛NAS",
"version": "1.3.6",
"version": "1.3.7",
"documentation": "https://github.com/anxms/fn_nas",
"dependencies": [],
"codeowners": ["@anxms"],

View File

@@ -293,87 +293,228 @@ class SystemManager:
return {}
async def get_vol_usage(self) -> dict:
"""获取 /vol* 开头的存储卷使用信息"""
"""获取 /vol* 开头的存储卷使用信息,避免唤醒休眠磁盘"""
try:
# 优先使用字节单位
df_output = await self.coordinator.run_command("df -B 1 /vol* 2>/dev/null")
if df_output:
return self.parse_df_bytes(df_output)
# 首先尝试智能检测活跃卷
active_vols = await self.check_active_volumes()
df_output = await self.coordinator.run_command("df -h /vol*")
if df_output:
return self.parse_df_human_readable(df_output)
if active_vols:
# 只查询活跃的卷,避免使用通配符可能唤醒所有磁盘
vol_list = " ".join(active_vols)
df_output = await self.coordinator.run_command(f"df -B 1 {vol_list} 2>/dev/null")
if df_output:
result = self.parse_df_bytes(df_output)
if result: # 确保有数据返回
return result
df_output = await self.coordinator.run_command(f"df -h {vol_list} 2>/dev/null")
if df_output:
result = self.parse_df_human_readable(df_output)
if result: # 确保有数据返回
return result
# 如果智能检测失败,回退到传统方法(仅在必要时)
self._debug_log("智能卷检测无结果,回退到传统检测方法")
# 优先使用字节单位,但添加错误处理
df_output = await self.coordinator.run_command("df -B 1 /vol* 2>/dev/null || true")
if df_output and "No such file or directory" not in df_output:
result = self.parse_df_bytes(df_output)
if result:
return result
df_output = await self.coordinator.run_command("df -h /vol* 2>/dev/null || true")
if df_output and "No such file or directory" not in df_output:
result = self.parse_df_human_readable(df_output)
if result:
return result
# 最后的回退:尝试检测任何挂载的卷
mount_output = await self.coordinator.run_command("mount | grep '/vol' || true")
if mount_output:
vol_points = []
for line in mount_output.splitlines():
parts = line.split()
for part in parts:
if part.startswith('/vol') and part not in vol_points:
vol_points.append(part)
if vol_points:
self._debug_log(f"从mount输出检测到卷: {vol_points}")
vol_list = " ".join(vol_points)
df_output = await self.coordinator.run_command(f"df -h {vol_list} 2>/dev/null || true")
if df_output:
return self.parse_df_human_readable(df_output)
self._debug_log("所有存储卷检测方法都失败,返回空字典")
return {}
except Exception as e:
self.logger.error("获取存储卷信息失败: %s", str(e))
self._error_log(f"获取存储卷信息失败: {str(e)}")
return {}
async def check_active_volumes(self) -> list:
"""检查当前活跃的存储卷,避免唤醒休眠磁盘"""
try:
# 获取所有挂载点,这个操作不会访问磁盘内容
mount_output = await self.coordinator.run_command("mount | grep '/vol' 2>/dev/null || true")
if not mount_output:
self._debug_log("未找到任何/vol挂载点")
return []
active_vols = []
for line in mount_output.splitlines():
if '/vol' in line:
# 提取挂载点
parts = line.split()
mount_point = None
# 查找挂载点(通常在 'on' 关键词之后)
try:
on_index = parts.index('on')
if on_index + 1 < len(parts):
candidate = parts[on_index + 1]
# 严格检查是否以/vol开头
if candidate.startswith('/vol'):
mount_point = candidate
except ValueError:
# 如果没有 'on' 关键词,查找以/vol开头的部分
for part in parts:
if part.startswith('/vol'):
mount_point = part
break
# 过滤挂载点:只保留根级别的/vol*挂载点
if mount_point and self.is_root_vol_mount(mount_point):
# 检查这个卷对应的磁盘是否活跃
is_active = await self.is_volume_disk_active(mount_point)
if is_active:
active_vols.append(mount_point)
self._debug_log(f"添加活跃卷: {mount_point}")
else:
# 即使磁盘不活跃,也添加到列表中,但标记为可能休眠
# 这样可以保证有基本的存储信息
active_vols.append(mount_point)
self._debug_log(f"{mount_point} 对应磁盘可能休眠,但仍包含在检测中")
else:
self._debug_log(f"跳过非根级别vol挂载点: {mount_point}")
# 去重并排序
active_vols = sorted(list(set(active_vols)))
self._debug_log(f"最终检测到的根级别/vol存储卷: {active_vols}")
return active_vols
except Exception as e:
self._debug_log(f"检查活跃存储卷失败: {e}")
return []
def is_root_vol_mount(self, mount_point: str) -> bool:
"""检查是否为根级别的/vol挂载点"""
if not mount_point or not mount_point.startswith('/vol'):
return False
# 移除开头的/vol部分进行分析
remainder = mount_point[4:] # 去掉'/vol'
# 如果remainder为空说明是/vol这是根级别
if not remainder:
return True
# 如果remainder只是数字如/vol1, /vol2这是根级别
if remainder.isdigit():
return True
# 如果remainder是单个字母或字母数字组合且没有斜杠也认为是根级别
# 例如:/vola, /volb, /vol1a 等
if '/' not in remainder and len(remainder) <= 3:
return True
# 其他情况都认为是子目录,如:
# /vol1/docker/overlay2/...
# /vol1/data/...
# /vol1/config/...
self._debug_log(f"检测到子目录挂载点: {mount_point}")
return False
def parse_df_bytes(self, df_output: str) -> dict:
"""解析df命令的字节输出"""
volumes = {}
for line in df_output.splitlines()[1:]:
parts = line.split()
if len(parts) < 6:
continue
try:
for line in df_output.splitlines()[1:]: # 跳过标题行
parts = line.split()
if len(parts) < 6:
continue
mount_point = parts[-1]
# 只处理 /vol 开头的挂载点
if not mount_point.startswith("/vol"):
continue
mount_point = parts[-1]
# 严格检查只处理根级别的 /vol 挂载点
if not self.is_root_vol_mount(mount_point):
self._debug_log(f"跳过非根级别vol挂载点: {mount_point}")
continue
try:
size_bytes = int(parts[1])
used_bytes = int(parts[2])
avail_bytes = int(parts[3])
use_percent = parts[4]
try:
size_bytes = int(parts[1])
used_bytes = int(parts[2])
avail_bytes = int(parts[3])
use_percent = parts[4]
def bytes_to_human(b):
for unit in ['', 'K', 'M', 'G', 'T']:
if abs(b) < 1024.0:
return f"{b:.1f}{unit}"
b /= 1024.0
return f"{b:.1f}P"
def bytes_to_human(b):
for unit in ['', 'K', 'M', 'G', 'T']:
if abs(b) < 1024.0:
return f"{b:.1f}{unit}"
b /= 1024.0
return f"{b:.1f}P"
volumes[mount_point] = {
"filesystem": parts[0],
"size": bytes_to_human(size_bytes),
"used": bytes_to_human(used_bytes),
"available": bytes_to_human(avail_bytes),
"use_percent": use_percent
}
except (ValueError, IndexError) as e:
self.logger.debug("解析存储卷行失败: %s - %s", line, str(e))
continue
volumes[mount_point] = {
"filesystem": parts[0],
"size": bytes_to_human(size_bytes),
"used": bytes_to_human(used_bytes),
"available": bytes_to_human(avail_bytes),
"use_percent": use_percent
}
self._debug_log(f"添加根级别/vol存储卷信息: {mount_point}")
except (ValueError, IndexError) as e:
self._debug_log(f"解析存储卷行失败: {line} - {str(e)}")
continue
except Exception as e:
self._error_log(f"解析df字节输出失败: {e}")
return volumes
def parse_df_human_readable(self, df_output: str) -> dict:
"""解析df命令输出"""
volumes = {}
for line in df_output.splitlines()[1:]:
parts = line.split()
if len(parts) < 6:
continue
try:
for line in df_output.splitlines()[1:]: # 跳过标题行
parts = line.split()
if len(parts) < 6:
continue
mount_point = parts[-1]
if not mount_point.startswith("/vol"):
continue
mount_point = parts[-1]
# 严格检查只处理根级别的 /vol 挂载点
if not self.is_root_vol_mount(mount_point):
self._debug_log(f"跳过非根级别vol挂载点: {mount_point}")
continue
try:
size = parts[1]
used = parts[2]
avail = parts[3]
use_percent = parts[4]
try:
size = parts[1]
used = parts[2]
avail = parts[3]
use_percent = parts[4]
volumes[mount_point] = {
"filesystem": parts[0],
"size": size,
"used": used,
"available": avail,
"use_percent": use_percent
}
except (ValueError, IndexError) as e:
self.logger.debug("解析存储卷行失败: %s - %s", line, str(e))
continue
volumes[mount_point] = {
"filesystem": parts[0],
"size": size,
"used": used,
"available": avail,
"use_percent": use_percent
}
self._debug_log(f"添加根级别/vol存储卷信息: {mount_point}")
except (ValueError, IndexError) as e:
self._debug_log(f"解析存储卷行失败: {line} - {str(e)}")
continue
except Exception as e:
self._error_log(f"解析df输出失败: {e}")
return volumes