S8ActiveFlowStuckScanJob.cs 5.1 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102
  1. using Admin.NET.Plugin.AiDOP.Service.S8;
  2. using Furion.Schedule;
  3. using Microsoft.Extensions.Configuration;
  4. using Microsoft.Extensions.DependencyInjection;
  5. using Microsoft.Extensions.Logging;
  6. namespace Admin.NET.Plugin.AiDOP.Job;
  7. /// <summary>
  8. /// 定期扫描挂有 ActiveFlow 且长时间未更新的 S8 异常,输出告警日志和通知记录。
  9. /// </summary>
  10. [JobDetail("job_s8_active_flow_stuck_scan", Description = "S8 ActiveFlow 卡死扫描",
  11. GroupName = "default", Concurrent = false)]
  12. [Period(300000, TriggerId = "trigger_s8_active_flow_stuck_scan", Description = "每5分钟执行", RunOnStart = true)]
  13. public class S8ActiveFlowStuckScanJob : IJob
  14. {
  15. private readonly IServiceScopeFactory _scopeFactory;
  16. private readonly IConfiguration _configuration;
  17. private readonly ILogger _logger;
  18. // S8-SCHEDULER-P0-BLEEDING-STOP-CONFIG-1:禁用日志单进程内只输出一次,避免每 tick 刷日志。
  19. // _firstParseFailLogged 控制 AIDOP_* env 解析失败 warn 单进程内只输出一次(首个失败 env 名)。
  20. private static int _firstDisabledLogged;
  21. private static int _firstParseFailLogged;
  22. // S8-SCHEDULER-CONFIG-ORDER-FIX(合入 P0):AIDOP_* env override 名称。
  23. // Furion 4.9.8.24 JSON 装配链后置导致 ASP.NET Core env vars 被 JSON 覆盖,
  24. // 故此处用 Environment.GetEnvironmentVariable 做 S8 调度开关专项 override。
  25. private const string EnvSchedulerEnabled = "AIDOP_SCHEDULER_ENABLED";
  26. private const string EnvS8SchedulerEnabled = "AIDOP_S8_SCHEDULER_ENABLED";
  27. // S8-ACTIVE-FLOW-TENANT-FILTER-P2-4:当前部署仅租户 1 / 工厂 1 投产,
  28. // stuck 候选与 dedup 查询补 tenant/factory 过滤;orphan 路径保持 global 0/0 不变。
  29. private const long DefaultTenantId = 1;
  30. private const long DefaultFactoryId = 1;
  31. public S8ActiveFlowStuckScanJob(IServiceScopeFactory scopeFactory, IConfiguration configuration, ILoggerFactory loggerFactory)
  32. {
  33. _scopeFactory = scopeFactory;
  34. _configuration = configuration;
  35. _logger = loggerFactory.CreateLogger(nameof(S8ActiveFlowStuckScanJob));
  36. }
  37. public async Task ExecuteAsync(JobExecutingContext context, CancellationToken stoppingToken)
  38. {
  39. // S8-SCHEDULER-P0-BLEEDING-STOP-CONFIG-1:环境级 + S8 业务级双开关 gate。
  40. // 任一为 false 时直接早退;早退前不访问任何业务 service / DB。
  41. // 注意:[Period(... RunOnStart = true)] 在启动期会立即触发一次,此处 gate 同样生效,
  42. // 确保 dev/未授权环境启动时不会立刻进入 ScanAsync。
  43. // env override(合入 CONFIG-ORDER-FIX):env 不存在/空白 → 沿用 JSON;
  44. // 存在但 bool.TryParse 失败 → 保留 JSON 值 + 单次 warn(首个失败 env 名,不输出 value)。
  45. var schedulerEnabled = _configuration.GetValue("Scheduler:Enabled", true);
  46. var s8SchedulerEnabled = _configuration.GetValue("S8:Scheduler:Enabled", false);
  47. string? parseFailEnvName = null;
  48. var envSchedulerRaw = Environment.GetEnvironmentVariable(EnvSchedulerEnabled);
  49. if (!string.IsNullOrWhiteSpace(envSchedulerRaw))
  50. {
  51. if (bool.TryParse(envSchedulerRaw, out var v)) schedulerEnabled = v;
  52. else parseFailEnvName ??= EnvSchedulerEnabled;
  53. }
  54. var envS8SchedulerRaw = Environment.GetEnvironmentVariable(EnvS8SchedulerEnabled);
  55. if (!string.IsNullOrWhiteSpace(envS8SchedulerRaw))
  56. {
  57. if (bool.TryParse(envS8SchedulerRaw, out var v)) s8SchedulerEnabled = v;
  58. else parseFailEnvName ??= EnvS8SchedulerEnabled;
  59. }
  60. if (parseFailEnvName != null && Interlocked.Exchange(ref _firstParseFailLogged, 1) == 0)
  61. {
  62. _logger.LogWarning(
  63. "S8ActiveFlowStuckScanJob env override 解析失败:{EnvName},已沿用配置值",
  64. parseFailEnvName);
  65. }
  66. if (!schedulerEnabled || !s8SchedulerEnabled)
  67. {
  68. if (Interlocked.Exchange(ref _firstDisabledLogged, 1) == 0)
  69. {
  70. _logger.LogInformation(
  71. "S8ActiveFlowStuckScanJob 被配置禁用:Scheduler:Enabled={Scheduler} S8:Scheduler:Enabled={S8Scheduler}",
  72. schedulerEnabled, s8SchedulerEnabled);
  73. }
  74. return;
  75. }
  76. using var scope = _scopeFactory.CreateScope();
  77. var watchService = scope.ServiceProvider.GetRequiredService<S8ActiveFlowWatchService>();
  78. try
  79. {
  80. var alertCount = await watchService.ScanAsync(DefaultTenantId, DefaultFactoryId, stoppingToken);
  81. if (alertCount > 0)
  82. _logger.LogInformation("S8ActiveFlowStuckScanJob 本轮新增 {Count} 条卡死告警", alertCount);
  83. }
  84. catch (OperationCanceledException) when (stoppingToken.IsCancellationRequested)
  85. {
  86. _logger.LogInformation("S8ActiveFlowStuckScanJob 收到停止信号,结束本轮扫描");
  87. }
  88. catch (Exception ex)
  89. {
  90. _logger.LogError(ex, "S8ActiveFlowStuckScanJob 扫描失败");
  91. }
  92. }
  93. }