diff --git a/TODO.md b/TODO.md index a4e1f72..9bf0d20 100644 --- a/TODO.md +++ b/TODO.md @@ -5,7 +5,7 @@ ## 测量 - 路由跳数、路径漂移和 trace -- 握手耗时：TCP 连接耗时，TLS / QUIC 握手耗时（QUIC 目前没有真实握手积木） +- QUIC 握手耗时：TCP 建连和 TLS 握手已经有了（`connect` 任务）。.NET 的 QUIC 依赖系统里的 msquic 原生库，和节点「只依赖 OpenSSL、NativeAOT 单文件」冲突，要做只能自己写最小的 QUIC Initial 握手探测 - 网页长截图 - pcap / eBPF 采集 @@ -13,7 +13,6 @@ - 单个 Agent 的任务（目前是线路模板加按 Agent 的覆盖，没有只属于某一个 Agent 的独立任务） - 原生 webhook 投递；告警（基于 webhook） -- Agent 和 Probe 的定时重启（可选，例如早上四点） ## 部署 diff --git a/backend/agent/src/MoShiro.Agent/AgentApp.cs b/backend/agent/src/MoShiro.Agent/AgentApp.cs index e8a497f..26e4435 100644 --- a/backend/agent/src/MoShiro.Agent/AgentApp.cs +++ b/backend/agent/src/MoShiro.Agent/AgentApp.cs @@ -16,6 +16,7 @@ public sealed class AgentApp : IDisposable private readonly AgentConfig _config; private readonly bool _ipv6Supported; + private readonly RestartController? _restart; private readonly LoadedNodeIdentity _identity; private readonly TimeProvider _time; private readonly INodeLog _log; @@ -37,8 +38,9 @@ public sealed class AgentApp : IDisposable private readonly FleetLink _link; /// 配置目录（容器里的 /etc/moshiro）：宿主机更新器和节点在这里交换清单、请求和状态。 - public AgentApp(AgentConfig config, LoadedNodeIdentity identity, string dataDirectory, TimeProvider time, INodeLog log) + public AgentApp(AgentConfig config, LoadedNodeIdentity identity, string dataDirectory, TimeProvider time, INodeLog log, RestartController? restart = null) { + _restart = restart; _update = new UpdateBridge(dataDirectory, new ConfigStore(dataDirectory).Write, log); _config = config; _identity = identity; @@ -219,6 +221,19 @@ public sealed class AgentApp : IDisposable }); } + if (Allow(MoShiroMethods.Schedule) && _restart is not null) + { + router.OnNotify(MoShiroMethods.Schedule, MoShiroJsonContext.Default.ScheduleParams, (_, schedule, _) => + { + if (!_restart.SetRemote(schedule.RestartAt)) + { + _log.Warn("ignoring an invalid restart schedule from the panel"); + } + + return ValueTask.CompletedTask; + }); + } + if (Allow(MoShiroMethods.UpdateApply)) { router.OnNotify(MoShiroMethods.UpdateApply, MoShiroJsonContext.Default.UpdateOfferParams, (_, _, _) => diff --git a/backend/agent/src/MoShiro.Agent/AgentConfig.cs b/backend/agent/src/MoShiro.Agent/AgentConfig.cs index 1331b8d..9032de6 100644 --- a/backend/agent/src/MoShiro.Agent/AgentConfig.cs +++ b/backend/agent/src/MoShiro.Agent/AgentConfig.cs @@ -22,6 +22,10 @@ public sealed class AgentConfig [JsonPropertyName("reporting")] public ReportingSection Reporting { get; set; } = new(); + + /// 每天定时重启一次，例如 04:00+08:00（不写偏移按 UTC）；不写就不重启。环境变量 MOSHIRO_RESTART_AT 优先。 + [JsonPropertyName("restart_at")] + public string? RestartAt { get; set; } } public sealed class ReportingSection diff --git a/backend/agent/src/MoShiro.Agent/Program.cs b/backend/agent/src/MoShiro.Agent/Program.cs index 096bf85..493a37b 100644 --- a/backend/agent/src/MoShiro.Agent/Program.cs +++ b/backend/agent/src/MoShiro.Agent/Program.cs @@ -22,9 +22,11 @@ for (int i = 0; i < args.Length - 1; i++) AgentConfig config; LoadedNodeIdentity identity; +RestartController restart; try { config = NodeConfig.Load(configPath, AgentConfigContext.Default.AgentConfig); + restart = new RestartController(RestartSchedule.Resolve(config.RestartAt)); identity = LoadedNodeIdentity.Load(config.Fleet.IdentityDir); } catch (Exception ex) when (ex is IOException or InvalidDataException or InvalidOperationException or System.Text.Json.JsonException or System.Security.Cryptography.CryptographicException) @@ -33,10 +35,11 @@ catch (Exception ex) when (ex is IOException or InvalidDataException or InvalidO return 1; } +using var restartScope = restart; using (identity) { - using var app = new AgentApp(config, identity, Path.GetDirectoryName(Path.GetFullPath(configPath))!, time, log); - await ServiceShell.RunAsync(app.RunAsync, log); + using var app = new AgentApp(config, identity, Path.GetDirectoryName(Path.GetFullPath(configPath))!, time, log, restart); + await ServiceShell.RunAsync(app.RunAsync, log, restart, time); } return 0; diff --git a/backend/fleet/src/Fleet.Hosting/Hosting.cs b/backend/fleet/src/Fleet.Hosting/Hosting.cs index 66faf23..ded35be 100644 --- a/backend/fleet/src/Fleet.Hosting/Hosting.cs +++ b/backend/fleet/src/Fleet.Hosting/Hosting.cs @@ -126,7 +126,8 @@ public static class PlatformCheck /// 处理 Ctrl+C 和 SIGTERM：取消令牌，让各个模块有序退出。 public static class ServiceShell { - public static async Task RunAsync(Func run, INodeLog log) + /// 每天定时有序退出一次（见 ），重新拉起由服务管理器负责。 + public static async Task RunAsync(Func run, INodeLog log, RestartController? restart = null, TimeProvider? time = null) { using var cts = new CancellationTokenSource(); using PosixSignalRegistration term = PosixSignalRegistration.Create(PosixSignal.SIGTERM, context => @@ -156,6 +157,7 @@ public static class ServiceShell cts.Cancel(); }); Task trimmer = TrimHeapAsync(cts.Token); + Task scheduled = restart is null ? Task.CompletedTask : WaitForRestartAsync(restart, time ?? TimeProvider.System, log, cts); try { await run(cts.Token); @@ -168,6 +170,38 @@ public static class ServiceShell { await cts.CancelAsync(); await trimmer; + await scheduled; + } + } + + private static async Task WaitForRestartAsync(RestartController restart, TimeProvider time, INodeLog log, CancellationTokenSource cts) + { + while (!cts.IsCancellationRequested) + { + // 先取令牌、再取设置：中间被改动的话令牌会被取消，下一轮重新排 + CancellationToken changed = restart.ChangeToken; + RestartSchedule? schedule = restart.Current; + using CancellationTokenSource wait = CancellationTokenSource.CreateLinkedTokenSource(cts.Token, changed); + try + { + if (schedule is null) + { + await Task.Delay(Timeout.Infinite, wait.Token); + continue; + } + + DateTimeOffset next = schedule.NextAfter(time.GetUtcNow() + RestartSchedule.MinimumUptime); + log.Info("restart scheduled", ("at", next.ToString("O", System.Globalization.CultureInfo.InvariantCulture)), ("source", restart.IsLocal ? "local" : "panel")); + await Task.Delay(next - time.GetUtcNow(), time, wait.Token); + } + catch (OperationCanceledException) + { + continue; // 进程在退出，或者设置被改了：循环条件和下一轮会处理 + } + + log.Info("stopping", ("result", "scheduled_restart")); + await cts.CancelAsync(); + return; } } diff --git a/backend/fleet/src/Fleet.Hosting/RestartController.cs b/backend/fleet/src/Fleet.Hosting/RestartController.cs new file mode 100644 index 0000000..a61fc36 --- /dev/null +++ b/backend/fleet/src/Fleet.Hosting/RestartController.cs @@ -0,0 +1,73 @@ +namespace Fleet.Hosting; + +/// +/// 当前生效的定时重启：节点本地的设置优先；本地没有写时用面板下发的，面板改了之后立即重新排下一次，不需要重启进程。 +/// +public sealed class RestartController : IDisposable +{ + private readonly RestartSchedule? _local; + private readonly object _gate = new(); + private RestartSchedule? _remote; + private CancellationTokenSource _changed = new(); + + public RestartController(RestartSchedule? local = null) + { + _local = local; + } + + /// 本地有设置时为 true：面板下发的不生效。 + public bool IsLocal => _local is not null; + + public RestartSchedule? Current + { + get + { + lock (_gate) + { + return _local ?? _remote; + } + } + } + + /// 当前设置被改动时会被取消的令牌；先取令牌、再取 ，就不会漏掉中间的改动。 + public CancellationToken ChangeToken + { + get + { + lock (_gate) + { + return _changed.Token; + } + } + } + + /// 面板下发的设置；为空表示面板关闭了定时重启。返回 false 表示文本不合法（节点忽略它）。 + public bool SetRemote(string? text) + { + RestartSchedule? schedule = null; + if (!string.IsNullOrWhiteSpace(text) && !RestartSchedule.TryParse(text, out schedule)) + { + return false; + } + + CancellationTokenSource previous; + lock (_gate) + { + _remote = schedule; + previous = _changed; + _changed = new CancellationTokenSource(); + } + + previous.Cancel(); + previous.Dispose(); + return true; + } + + public void Dispose() + { + lock (_gate) + { + _changed.Dispose(); + } + } +} diff --git a/backend/fleet/src/Fleet.Hosting/RestartSchedule.cs b/backend/fleet/src/Fleet.Hosting/RestartSchedule.cs new file mode 100644 index 0000000..1c730f4 --- /dev/null +++ b/backend/fleet/src/Fleet.Hosting/RestartSchedule.cs @@ -0,0 +1,105 @@ +using System.Globalization; + +namespace Fleet.Hosting; + +/// +/// 每天固定时刻有序退出一次，由 systemd、容器引擎或 procd 的「总是重启」拉起。写法 HH:mm（按 UTC）、 +/// HH:mm±hh:mm 或 HH:mm 加 Z：容器里没有时区数据库，所以偏移要写明，例如 04:00+08:00。 +/// 进程自己退出，不需要任何执行外部命令的能力。 +/// +public sealed class RestartSchedule +{ + /// 环境变量，设置了就覆盖配置文件里的值。 + public const string EnvironmentVariable = "MOSHIRO_RESTART_AT"; + + /// 刚启动的一段时间内不会触发：重启之后不会因为时刻还没过去而再重启一次，也避免重启风暴。 + public static readonly TimeSpan MinimumUptime = TimeSpan.FromMinutes(10); + + private readonly TimeSpan _time; + private readonly TimeSpan _offset; + + private RestartSchedule(TimeSpan time, TimeSpan offset) + { + _time = time; + _offset = offset; + } + + public static bool TryParse(string? text, out RestartSchedule? schedule) + { + schedule = null; + string value = text?.Trim() ?? ""; + // HH:mm，后面可以跟 Z 或 ±hh:mm + if (value.Length < 5 || value[2] != ':' || + !TryTwoDigits(value, 0, out int hour) || !TryTwoDigits(value, 3, out int minute) || hour > 23 || minute > 59) + { + return false; + } + + TimeSpan offset = TimeSpan.Zero; + string rest = value[5..]; + if (rest.Length > 0 && rest != "Z") + { + if (rest.Length != 6 || rest[0] is not ('+' or '-') || rest[3] != ':' || + !TryTwoDigits(rest, 1, out int offsetHours) || !TryTwoDigits(rest, 4, out int offsetMinutes) || offsetMinutes > 59) + { + return false; + } + + offset = new TimeSpan(offsetHours, offsetMinutes, 0); + if (offset > TimeSpan.FromHours(14)) + { + return false; + } + + if (rest[0] == '-') + { + offset = -offset; + } + } + + schedule = new RestartSchedule(new TimeSpan(hour, minute, 0), offset); + return true; + } + + private static bool TryTwoDigits(string text, int start, out int value) + { + value = 0; + if (!char.IsAsciiDigit(text[start]) || !char.IsAsciiDigit(text[start + 1])) + { + return false; + } + + value = (text[start] - '0') * 10 + (text[start + 1] - '0'); + return true; + } + + /// + /// 环境变量优先，其次配置文件的值；都没有就不定时重启。写错了抛 ， + /// 启动时就能看到，而不是悄悄不生效。 + /// + public static RestartSchedule? Resolve(string? configured) + { + string? text = Environment.GetEnvironmentVariable(EnvironmentVariable); + if (string.IsNullOrWhiteSpace(text)) + { + text = configured; + } + + if (string.IsNullOrWhiteSpace(text)) + { + return null; + } + + return TryParse(text, out RestartSchedule? schedule) + ? schedule + : throw new InvalidOperationException($"restart_at must look like 04:00, 04:00Z or 04:00+08:00 (got '{text}')"); + } + + /// 严格晚于 的下一个触发时刻。 + public DateTimeOffset NextAfter(DateTimeOffset now) + { + DateTimeOffset local = now.ToOffset(_offset); + DateTimeOffset candidate = new DateTimeOffset(local.Year, local.Month, local.Day, 0, 0, 0, _offset) + _time; + return candidate <= local ? candidate.AddDays(1) : candidate; + } +} diff --git a/backend/fleet/tests/Fleet.Tests/Fleet.Tests.csproj b/backend/fleet/tests/Fleet.Tests/Fleet.Tests.csproj index 146ccda..336bf03 100644 --- a/backend/fleet/tests/Fleet.Tests/Fleet.Tests.csproj +++ b/backend/fleet/tests/Fleet.Tests/Fleet.Tests.csproj @@ -9,6 +9,7 @@ + diff --git a/backend/fleet/tests/Fleet.Tests/RestartScheduleTests.cs b/backend/fleet/tests/Fleet.Tests/RestartScheduleTests.cs new file mode 100644 index 0000000..84c8855 --- /dev/null +++ b/backend/fleet/tests/Fleet.Tests/RestartScheduleTests.cs @@ -0,0 +1,260 @@ +using Fleet.Core; +using Fleet.Hosting; + +namespace Fleet.Tests; + +public sealed class RestartScheduleTests +{ + private static RestartSchedule Parse(string text) + { + Assert.True(RestartSchedule.TryParse(text, out RestartSchedule? schedule), text); + return schedule!; + } + + [Theory] + [InlineData("04:00")] + [InlineData("04:00Z")] + [InlineData("04:00+08:00")] + [InlineData("23:59-05:30")] + [InlineData(" 00:00 ")] + public void Parse_AcceptsTimeWithOptionalOffset(string text) => Parse(text); + + [Theory] + [InlineData("")] + [InlineData("4:00")] + [InlineData("24:00")] + [InlineData("04:60")] + [InlineData("04:00+8:00")] + [InlineData("04:00+15:00")] + [InlineData("04:00 UTC")] + [InlineData("0400")] + [InlineData("04:00+08:60")] + public void Parse_RejectsEverythingElse(string text) => Assert.False(RestartSchedule.TryParse(text, out _)); + + [Fact] + public void NextAfter_IsStrictlyLater_AndRollsToTomorrow() + { + RestartSchedule utc = Parse("04:00"); + Assert.Equal(new DateTimeOffset(2026, 10, 7, 4, 0, 0, TimeSpan.Zero), utc.NextAfter(new DateTimeOffset(2026, 10, 7, 3, 59, 59, TimeSpan.Zero))); + Assert.Equal(new DateTimeOffset(2026, 10, 8, 4, 0, 0, TimeSpan.Zero), utc.NextAfter(new DateTimeOffset(2026, 10, 7, 4, 0, 0, TimeSpan.Zero))); + Assert.Equal(new DateTimeOffset(2026, 10, 8, 4, 0, 0, TimeSpan.Zero), utc.NextAfter(new DateTimeOffset(2026, 10, 7, 20, 0, 0, TimeSpan.Zero))); + } + + [Fact] + public void NextAfter_UsesTheWrittenOffset() + { + // 04:00+08:00 就是 UTC 前一天 20:00 + RestartSchedule china = Parse("04:00+08:00"); + Assert.Equal(new DateTimeOffset(2026, 10, 6, 20, 0, 0, TimeSpan.Zero), china.NextAfter(new DateTimeOffset(2026, 10, 6, 19, 0, 0, TimeSpan.Zero))); + Assert.Equal(new DateTimeOffset(2026, 10, 7, 20, 0, 0, TimeSpan.Zero), china.NextAfter(new DateTimeOffset(2026, 10, 6, 20, 0, 0, TimeSpan.Zero))); + } + + [Fact] + public void NextAfter_WithMinimumUptime_DoesNotRestartAgainRightAfterARestart() + { + RestartSchedule schedule = Parse("04:00"); + // 04:00:03 重启后启动：加上最短运行时间再算，下一次是明天，而不是几秒后 + var started = new DateTimeOffset(2026, 10, 7, 4, 0, 3, TimeSpan.Zero); + Assert.Equal(new DateTimeOffset(2026, 10, 8, 4, 0, 0, TimeSpan.Zero), schedule.NextAfter(started + RestartSchedule.MinimumUptime)); + // 03:55 启动：离 04:00 只有 5 分钟，不够最短运行时间，顺延到明天 + var early = new DateTimeOffset(2026, 10, 7, 3, 55, 0, TimeSpan.Zero); + Assert.Equal(new DateTimeOffset(2026, 10, 8, 4, 0, 0, TimeSpan.Zero), schedule.NextAfter(early + RestartSchedule.MinimumUptime)); + } + + [Fact] + public void Resolve_ReportsABadValueInsteadOfIgnoringIt() + { + Assert.Null(RestartSchedule.Resolve(null)); + Assert.Null(RestartSchedule.Resolve(" ")); + InvalidOperationException error = Assert.Throws(() => RestartSchedule.Resolve("tomorrow")); + Assert.Contains("restart_at", error.Message); + } + + [Fact] + public async Task ServiceShell_StopsAtTheScheduledTime() + { + var time = new ImmediateTimeProvider(new DateTimeOffset(2026, 10, 7, 12, 0, 0, TimeSpan.Zero)); + var log = new ListLog(); + bool stoppedByToken = false; + + // 业务循环只在被取消时返回；定时重启到点以后令牌被取消，RunAsync 正常返回 + await ServiceShell.RunAsync(async ct => + { + try + { + await Task.Delay(Timeout.Infinite, ct); + } + catch (OperationCanceledException) + { + stoppedByToken = true; + throw; + } + }, log, new RestartController(Parse("04:00")), time).WaitAsync(TimeSpan.FromSeconds(10)); + + Assert.True(stoppedByToken); + Assert.Contains(log.Lines, l => l.Contains("scheduled_restart", StringComparison.Ordinal)); + Assert.Contains(log.Lines, l => l.Contains("restart scheduled", StringComparison.Ordinal)); + } + + [Fact] + public async Task ServiceShell_PicksUpAScheduleThePanelSendsLater() + { + var time = new ImmediateTimeProvider(new DateTimeOffset(2026, 10, 7, 12, 0, 0, TimeSpan.Zero)); + var log = new ListLog(); + using var controller = new RestartController(); + + Task shell = ServiceShell.RunAsync(ct => Task.Delay(Timeout.Infinite, ct), log, controller, time); + await Task.Delay(300); + Assert.False(shell.IsCompleted, "no schedule yet: the node keeps running"); + + Assert.True(controller.SetRemote("04:30+08:00")); + await shell.WaitAsync(TimeSpan.FromSeconds(10)); + Assert.Contains(log.Lines, l => l.Contains("scheduled_restart", StringComparison.Ordinal)); + Assert.Contains(log.Lines, l => l.Contains("source=panel", StringComparison.Ordinal)); + } + + [Fact] + public async Task ServiceShell_ClearingTheScheduleCancelsTheRestart() + { + var time = new ManualTimeProvider(new DateTimeOffset(2026, 10, 7, 12, 0, 0, TimeSpan.Zero)); + var log = new ListLog(); + using var controller = new RestartController(); + + controller.SetRemote("04:30+08:00"); + + Task shell = ServiceShell.RunAsync(ct => Task.Delay(Timeout.Infinite, ct), log, controller, time); + await Task.Delay(300); + controller.SetRemote(null); // 面板关掉了 + await Task.Delay(300); + Assert.False(shell.IsCompleted); + Assert.Contains(log.Lines, l => l.Contains("restart scheduled", StringComparison.Ordinal)); + Assert.DoesNotContain(log.Lines, l => l.Contains("scheduled_restart", StringComparison.Ordinal)); + // 清理：把进程停掉 + Assert.True(controller.SetRemote("04:30+08:00")); + time.Advance(); + await shell.WaitAsync(TimeSpan.FromSeconds(10)); + } + + [Fact] + public void Controller_LocalSettingWinsOverThePanel_AndBadTextIsRejected() + { + using var controller = new RestartController(Parse("04:00")); + Assert.True(controller.IsLocal); + Assert.True(controller.SetRemote("05:00")); + Assert.Equal(new DateTimeOffset(2026, 10, 7, 4, 0, 0, TimeSpan.Zero), controller.Current!.NextAfter(new DateTimeOffset(2026, 10, 7, 0, 0, 0, TimeSpan.Zero))); + + using var remoteOnly = new RestartController(); + Assert.Null(remoteOnly.Current); + Assert.False(remoteOnly.SetRemote("whenever")); + Assert.Null(remoteOnly.Current); + Assert.True(remoteOnly.SetRemote("05:00")); + Assert.NotNull(remoteOnly.Current); + Assert.True(remoteOnly.SetRemote(null)); + Assert.Null(remoteOnly.Current); + } + + /// 计时器马上触发：用来验证「到点」之后的行为，不用真的等。 + private sealed class ImmediateTimeProvider : TimeProvider + { + private readonly DateTimeOffset _now; + + public ImmediateTimeProvider(DateTimeOffset now) => _now = now; + + public override DateTimeOffset GetUtcNow() => _now; + + public override ITimer CreateTimer(TimerCallback callback, object? state, TimeSpan dueTime, TimeSpan period) + { + // TrimHeapAsync 这类每 30 秒一次的循环也会走到这里；只让「很长」的延迟（定时重启，至少 10 分钟）马上触发 + var timer = new ImmediateTimer(callback, state); + if (dueTime >= RestartSchedule.MinimumUptime) + { + ThreadPool.QueueUserWorkItem(_ => timer.Fire()); + } + + return timer; + } + } + + /// 计时器只在 时触发：用来验证「还没到点」时不会退出。 + private sealed class ManualTimeProvider : TimeProvider + { + private readonly DateTimeOffset _now; + private readonly List _timers = []; + + public ManualTimeProvider(DateTimeOffset now) => _now = now; + + public override DateTimeOffset GetUtcNow() => _now; + + public override ITimer CreateTimer(TimerCallback callback, object? state, TimeSpan dueTime, TimeSpan period) + { + var timer = new ImmediateTimer(callback, state); + lock (_timers) + { + _timers.Add(timer); + } + + return timer; + } + + public void Advance() + { + ImmediateTimer[] due; + lock (_timers) + { + due = [.. _timers]; + _timers.Clear(); + } + + foreach (ImmediateTimer timer in due) + { + timer.Fire(); + } + } + } + + private sealed class ImmediateTimer : ITimer + { + private readonly TimerCallback _callback; + private readonly object? _state; + + public ImmediateTimer(TimerCallback callback, object? state) + { + _callback = callback; + _state = state; + } + + public void Fire() => _callback(_state); + + public bool Change(TimeSpan dueTime, TimeSpan period) => true; + + public void Dispose() + { + } + + public ValueTask DisposeAsync() => ValueTask.CompletedTask; + } + + private sealed class ListLog : INodeLog + { + private readonly List _lines = []; + + public IReadOnlyList Lines + { + get + { + lock (_lines) + { + return [.. _lines]; + } + } + } + + public void Write(NodeLogLevel level, string message, params (string Key, object? Value)[] fields) + { + lock (_lines) + { + _lines.Add(message + " " + string.Join(' ', fields.Select(f => f.Key + "=" + f.Value))); + } + } + } +} diff --git a/backend/probe/src/MoShiro.Probe/ProbeApp.cs b/backend/probe/src/MoShiro.Probe/ProbeApp.cs index 15d28df..f3eba1a 100644 --- a/backend/probe/src/MoShiro.Probe/ProbeApp.cs +++ b/backend/probe/src/MoShiro.Probe/ProbeApp.cs @@ -23,7 +23,7 @@ public sealed class ProbeApp : IDisposable private readonly INodeLog _log; private readonly FleetLink _link; - public ProbeApp(ProbeFile file, string configDirectory, TimeProvider time, INodeLog log) + public ProbeApp(ProbeFile file, string configDirectory, TimeProvider time, INodeLog log, RestartController? restart = null) { _log = log; var directory = new AgentDirectory(new ConfigStore(configDirectory), configDirectory, log); @@ -59,6 +59,15 @@ public sealed class ProbeApp : IDisposable .OnNotify(MoShiroMethods.UpdateApply, MoShiroJsonContext.Default.UpdateOfferParams, (_, _, _) => { update.RequestApply(); + return ValueTask.CompletedTask; + }) + .OnNotify(MoShiroMethods.Schedule, MoShiroJsonContext.Default.ScheduleParams, (_, schedule, _) => + { + if (restart is not null && !restart.SetRemote(schedule.RestartAt)) + { + log.Warn("ignoring an invalid restart schedule from the panel"); + } + return ValueTask.CompletedTask; }); _link = new FleetLink(options, router, [list], new Outbox(1 << 20), time, log); diff --git a/backend/probe/src/MoShiro.Probe/ProbeConfig.cs b/backend/probe/src/MoShiro.Probe/ProbeConfig.cs index 1057d90..5a3bc19 100644 --- a/backend/probe/src/MoShiro.Probe/ProbeConfig.cs +++ b/backend/probe/src/MoShiro.Probe/ProbeConfig.cs @@ -20,6 +20,10 @@ public sealed class ProbeFile [JsonPropertyName("server")] public string Server { get; set; } = ""; + /// 每天定时重启一次，例如 04:00+08:00（不写偏移按 UTC）；不写就不重启。环境变量 MOSHIRO_RESTART_AT 优先；这里是本地配置，从面板重新下载 probe.json 会丢掉它，所以容器里推荐用环境变量。 + [JsonPropertyName("restart_at")] + public string? RestartAt { get; set; } + [JsonPropertyName("limits")] public ProbeLimits Limits { get; set; } = new(); } diff --git a/backend/probe/src/MoShiro.Probe/Program.cs b/backend/probe/src/MoShiro.Probe/Program.cs index cadfb11..f463996 100644 --- a/backend/probe/src/MoShiro.Probe/Program.cs +++ b/backend/probe/src/MoShiro.Probe/Program.cs @@ -21,9 +21,11 @@ for (int i = 0; i < args.Length - 1; i++) } ProbeFile file; +RestartController restart; try { file = NodeConfig.Load(configPath, ProbeConfigContext.Default.ProbeFile); + restart = new RestartController(RestartSchedule.Resolve(file.RestartAt)); if (file.ProbeId.Length == 0 || file.PrivateKey.Length == 0 || !Uri.TryCreate(file.Server, UriKind.Absolute, out _)) { throw new InvalidOperationException("probe.json needs probe_id, private_key and server (download it from the panel: Probes → Create)"); @@ -35,11 +37,12 @@ catch (Exception ex) when (ex is IOException or InvalidDataException or InvalidO return 1; } +using var restartScope = restart; string directory = Path.GetDirectoryName(Path.GetFullPath(configPath))!; try { - using var app = new ProbeApp(file, directory, time, log); - await ServiceShell.RunAsync(app.RunAsync, log); + using var app = new ProbeApp(file, directory, time, log, restart); + await ServiceShell.RunAsync(app.RunAsync, log, restart, time); } catch (Exception ex) when (ex is FormatException or System.Security.Cryptography.CryptographicException) { diff --git a/backend/shared/MoShiro.Protocol/Rpc/MoShiroRpc.cs b/backend/shared/MoShiro.Protocol/Rpc/MoShiroRpc.cs index 5907bd6..b9f6da2 100644 --- a/backend/shared/MoShiro.Protocol/Rpc/MoShiroRpc.cs +++ b/backend/shared/MoShiro.Protocol/Rpc/MoShiroRpc.cs @@ -15,6 +15,7 @@ public static class MoShiroMethods public const string Host = "moshiro.host"; public const string UpdateOffer = "moshiro.update.offer"; public const string UpdateApply = "moshiro.update.apply"; + public const string Schedule = "moshiro.schedule"; // Probe ↔ Agent（点对点会话） public const string Clock = "moshiro.clock"; @@ -238,6 +239,17 @@ public sealed class UpdateOfferParams public string Signature { get; set; } = ""; } +/// +/// 面板给节点的定时重启设置（moshiro.schedule 通知）：restart_at 形如 04:30+08:00，为空表示不定时重启。 +/// 节点连上时补发、设置变化时群发；节点本地写了 restart_at 或环境变量 MOSHIRO_RESTART_AT 时以本地为准。 +/// 节点只是到点自己有序退出，由服务管理器拉起，不执行任何命令。 +/// +public sealed class ScheduleParams +{ + [JsonPropertyName("restart_at")] + public string? RestartAt { get; set; } +} + /// 宿主机更新器已经验证过的可用更新。 public sealed class UpdateAvailable { @@ -560,6 +572,7 @@ public static class FragmentKeys [JsonSerializable(typeof(AgentEventParams))] [JsonSerializable(typeof(ProbeAgentList))] [JsonSerializable(typeof(UpdateOfferParams))] +[JsonSerializable(typeof(ScheduleParams))] [JsonSerializable(typeof(UpdateStatus))] [JsonSerializable(typeof(ProbeHeartbeatExtra))] [JsonSerializable(typeof(ProbeHelloMeta))] diff --git a/backend/src/MoShiro.Backend.Api/Admin/SystemEndpoints.cs b/backend/src/MoShiro.Backend.Api/Admin/SystemEndpoints.cs index 00a6037..8c67d36 100644 --- a/backend/src/MoShiro.Backend.Api/Admin/SystemEndpoints.cs +++ b/backend/src/MoShiro.Backend.Api/Admin/SystemEndpoints.cs @@ -63,9 +63,14 @@ public static class SystemEndpoints private static async Task ListSettingsAsync(SettingsService settings, CancellationToken ct) => TypedResults.Json(new PageDto([.. (await settings.ListAsync(ct)).Select(Mappers.ToDto)], null), ApiJsonContext.Default.PageDtoSettingDto); - private static async Task PatchSettingsAsync(HttpContext http, SettingsService settings, AgentService agents, Dictionary changes) + private static async Task PatchSettingsAsync(HttpContext http, SettingsService settings, AgentService agents, ScheduleService schedule, Dictionary changes) { Result> result = await settings.PatchAsync(changes, http.RequestAborted); + if (result.IsSuccess && changes.Keys.Any(k => k.StartsWith("fleet.restart_", StringComparison.Ordinal))) + { + await schedule.SendToAllAsync(http.RequestAborted); // 定时重启设置变了：马上通知所有在线节点 + } + if (result.IsSuccess && changes.ContainsKey(SettingsService.AutoLink)) { await agents.AfterAgentSetChangedAsync(http.RequestAborted); // 自动连接开关变了：重新下发各 Probe 的 Agent 列表和各 Agent 允许的 Probe diff --git a/backend/src/MoShiro.Backend.Application/Admin/SettingsService.cs b/backend/src/MoShiro.Backend.Application/Admin/SettingsService.cs index 1577ec3..62102ee 100644 --- a/backend/src/MoShiro.Backend.Application/Admin/SettingsService.cs +++ b/backend/src/MoShiro.Backend.Application/Admin/SettingsService.cs @@ -14,6 +14,10 @@ public sealed class SettingsService public const string SitePublic = "site.public"; public const string SiteShowBuild = "site.show_build"; public const string AutoLink = "fleet.auto_link"; + public const string RestartEnabled = "fleet.restart_enabled"; + public const string RestartHour = "fleet.restart_hour"; + public const string RestartMinute = "fleet.restart_minute"; + public const string RestartOffsetMinutes = "fleet.restart_utc_offset_minutes"; public const string RawDays = "retention.raw_days"; public const string TestRawDays = "retention.test_raw_days"; public const string Metric10sHours = "retention.metric_10s_hours"; @@ -44,6 +48,10 @@ public sealed class SettingsService new(SitePublic, "bool", "false", 0, 1, "Whether anonymous visitors can read the public API."), new(SiteShowBuild, "bool", "false", 0, 1, "Whether the public pages show the versions, tags and build times of the backend, agents and probes."), new(AutoLink, "bool", "true", 0, 1, "Whether every probe connects to every agent automatically; when off, only the links chosen by hand are connected."), + new(RestartEnabled, "bool", "true", 0, 1, "Whether agents and probes restart themselves once a day (they exit in order and the service manager starts them again); a local restart_at on the node wins."), + new(RestartHour, "int", "4", 0, 23, "Hour of the daily restart, in the time zone below."), + new(RestartMinute, "int", "30", 0, 59, "Minute of the daily restart."), + new(RestartOffsetMinutes, "int", "480", -840, 840, "Time zone of the daily restart as minutes east of UTC (480 = Beijing)."), new(RawDays, "int", o.RawRetentionDays.ToString(CultureInfo.InvariantCulture), 1, 365, "Days to keep raw chunks of continuous tasks."), new(TestRawDays, "int", o.TestRawRetentionDays.ToString(CultureInfo.InvariantCulture), 1, 365, "Days to keep raw chunks of on-demand tests."), new(Metric10sHours, "int", o.Metric10sRetentionHours.ToString(CultureInfo.InvariantCulture), 1, 720, "Hours to keep the 10 s metric layer."), @@ -121,6 +129,12 @@ public sealed class SettingsService /// 访客能不能看到版本、tag 和构建时间（站点公开之后的另一道开关，默认关）。 public async ValueTask IsShowBuildAsync(CancellationToken ct) => await _store.GetAsync(SiteShowBuild, ct) == "true"; + public async ValueTask GetBoolAsync(string key, CancellationToken ct) + { + SettingSpec spec = Specs.First(s => s.Key == key); + return (await _store.GetAsync(key, ct) ?? spec.Default) == "true"; + } + public async ValueTask GetIntAsync(string key, CancellationToken ct) { SettingSpec spec = Specs.First(s => s.Key == key); diff --git a/backend/src/MoShiro.Backend.Application/Agents/MoShiroFleetHooks.cs b/backend/src/MoShiro.Backend.Application/Agents/MoShiroFleetHooks.cs index a2c6312..7c51ac3 100644 --- a/backend/src/MoShiro.Backend.Application/Agents/MoShiroFleetHooks.cs +++ b/backend/src/MoShiro.Backend.Application/Agents/MoShiroFleetHooks.cs @@ -53,11 +53,13 @@ public sealed class MoShiroNodeEventSink : INodeEventSink private readonly AgentService _service; private readonly NodeInfoService _nodes; private readonly ReleaseService _releases; + private readonly ScheduleService _schedule; private readonly IProbeStore _probes; - public MoShiroNodeEventSink(IAgentStore agents, AgentRuntime runtime, IEventStore events, LiveHub live, AgentService service, NodeInfoService nodes, ReleaseService releases, IProbeStore probes) + public MoShiroNodeEventSink(IAgentStore agents, AgentRuntime runtime, IEventStore events, LiveHub live, AgentService service, NodeInfoService nodes, ReleaseService releases, ScheduleService schedule, IProbeStore probes) { _releases = releases; + _schedule = schedule; _probes = probes; _nodes = nodes; _agents = agents; @@ -144,6 +146,7 @@ public sealed class MoShiroNodeEventSink : INodeEventSink } await _releases.OfferToNodeAsync(id, ct); + await _schedule.SendToNodeAsync(id, ct); break; case NodeEventKind.Offline: _live.Publish(new NodeSignal("probe", id.Value, false, nodeEvent.AtMs)); @@ -214,6 +217,7 @@ public sealed class MoShiroNodeEventSink : INodeEventSink AgentRecord? after = await _agents.GetAsync(id, ct); _nodes.ObserveAgentInBackground(id.Value, after?.PublicHost, meta.Platform); await _releases.OfferToNodeAsync(id, ct); + await _schedule.SendToNodeAsync(id, ct); if (before is not null && after is not null && (!before.TlsSpki.AsSpan().SequenceEqual(after.TlsSpki) || before.ProbeTcpPort != after.ProbeTcpPort || !before.ProbeUdpPorts.SequenceEqual(after.ProbeUdpPorts) || before.Ipv6Supported != after.Ipv6Supported) && after.Status == NodeStatus.Active) diff --git a/backend/src/MoShiro.Backend.Application/Agents/ScheduleService.cs b/backend/src/MoShiro.Backend.Application/Agents/ScheduleService.cs new file mode 100644 index 0000000..b71c197 --- /dev/null +++ b/backend/src/MoShiro.Backend.Application/Agents/ScheduleService.cs @@ -0,0 +1,54 @@ +using System.Globalization; +using Fleet.Core; + +namespace MoShiro.Backend.Application; + +/// +/// 节点的定时重启：设置在管理端（开关、时刻、时区偏移），节点连上时补发、设置变化时群发给所有在线节点。 +/// 面板只告诉节点「几点退出」，节点自己到点有序退出、由服务管理器拉起；节点本地写了设置时以本地为准。 +/// +public sealed class ScheduleService +{ + private readonly SettingsService _settings; + private readonly IAgentGateway _gateway; + + public ScheduleService(SettingsService settings, IAgentGateway gateway) + { + _settings = settings; + _gateway = gateway; + } + + /// 形如 04:30+08:00；开关关着时为空。 + public async ValueTask RestartAtAsync(CancellationToken ct) + { + if (!await _settings.GetBoolAsync(SettingsService.RestartEnabled, ct)) + { + return null; + } + + return Format( + await _settings.GetIntAsync(SettingsService.RestartHour, ct), + await _settings.GetIntAsync(SettingsService.RestartMinute, ct), + await _settings.GetIntAsync(SettingsService.RestartOffsetMinutes, ct)); + } + + public static string Format(int hour, int minute, int offsetMinutes) + { + int abs = Math.Abs(offsetMinutes); + return string.Create(CultureInfo.InvariantCulture, $"{hour:00}:{minute:00}{(offsetMinutes < 0 ? '-' : '+')}{abs / 60:00}:{abs % 60:00}"); + } + + /// 节点刚连上：补发当前设置（节点离线时改的设置也不会错过）。 + public async ValueTask SendToNodeAsync(NodeId node, CancellationToken ct) => + await _gateway.NotifyScheduleAsync(node, await RestartAtAsync(ct), ct); + + /// 管理端改了设置：通知所有在线节点。 + public async ValueTask SendToAllAsync(CancellationToken ct) + { + string? restartAt = await RestartAtAsync(ct); + foreach (NodeId node in _gateway.OnlineNodes) + { + await _gateway.NotifyScheduleAsync(node, restartAt, ct); + } + } +} diff --git a/backend/src/MoShiro.Backend.Application/ApplicationServiceCollectionExtensions.cs b/backend/src/MoShiro.Backend.Application/ApplicationServiceCollectionExtensions.cs index 70eb75b..df33d2b 100644 --- a/backend/src/MoShiro.Backend.Application/ApplicationServiceCollectionExtensions.cs +++ b/backend/src/MoShiro.Backend.Application/ApplicationServiceCollectionExtensions.cs @@ -37,6 +37,7 @@ public static class ApplicationServiceCollectionExtensions services.AddSingleton(); services.AddSingleton(); services.AddSingleton(); + services.AddSingleton(); services.AddSingleton(); services.AddHostedService(p => p.GetRequiredService()); services.AddSingleton(); diff --git a/backend/src/MoShiro.Backend.Application/Ports/IAgentGateway.cs b/backend/src/MoShiro.Backend.Application/Ports/IAgentGateway.cs index ed855df..b330cff 100644 --- a/backend/src/MoShiro.Backend.Application/Ports/IAgentGateway.cs +++ b/backend/src/MoShiro.Backend.Application/Ports/IAgentGateway.cs @@ -33,6 +33,9 @@ public interface IAgentGateway /// 把发布清单和签名原样交给节点（通知，不等回复）。 ValueTask NotifyUpdateOfferAsync(NodeId node, string manifest, string signature, CancellationToken ct); + /// 把定时重启设置交给节点（通知，不等回复）； 为空表示关闭。 + ValueTask NotifyScheduleAsync(NodeId node, string? restartAt, CancellationToken ct); + /// 请求节点更新到它手里的清单。节点不在线返回 false。 ValueTask NotifyUpdateApplyAsync(NodeId node, CancellationToken ct); } diff --git a/backend/src/MoShiro.Backend.Application/Tasks/TaskService.cs b/backend/src/MoShiro.Backend.Application/Tasks/TaskService.cs index 6bafbfd..50687fe 100644 --- a/backend/src/MoShiro.Backend.Application/Tasks/TaskService.cs +++ b/backend/src/MoShiro.Backend.Application/Tasks/TaskService.cs @@ -40,23 +40,35 @@ public sealed class TaskService public ValueTask GetAsync(TaskId id, CancellationToken ct) => _tasks.GetAsync(id, ct); - /// 首次启动时创建默认的线路任务模板（udp_echo、tcp_echo 持续运行），这样新 Agent 注册后自动获得任务。 + /// + /// 创建默认的线路任务模板，这样新 Agent 注册后自动获得任务：udp_echo、tcp_echo 持续运行，connect（TCP 建连和 TLS 握手耗时， + /// 每分钟 6 次）也持续运行。每一批默认任务只种一次，用设置里的标记记住；你删掉的默认模板不会被重新建出来， + /// 后来新增的默认任务（connect）在已有部署升级后会补建一次。 + /// public async ValueTask SeedDefaultsAsync(CancellationToken ct) { - if (await _settings.GetAsync("tasks.seeded", ct) is not null) + if (await _settings.GetAsync("tasks.seeded", ct) is null) { - return; + await SeedAsync([Kinds.UdpEcho, Kinds.TcpEcho], ct); + await _settings.SetAsync("tasks.seeded", "1", ct); } + if (await _settings.GetAsync("tasks.seeded.connect", ct) is null) + { + await SeedAsync([Kinds.Connect], ct); + await _settings.SetAsync("tasks.seeded.connect", "1", ct); + } + } + + private async ValueTask SeedAsync(KindSpec[] kinds, CancellationToken ct) + { long now = _time.GetUtcNow().ToUnixTimeMilliseconds(); - foreach (KindSpec kind in new[] { Kinds.UdpEcho, Kinds.TcpEcho }) + foreach (KindSpec kind in kinds) { await _tasks.CreateAsync( new TaskRecord(new TaskId(Ulid.NewPrefixed(TaskId.Prefix, _time)), EffectiveTasks.LineScope, null, null, kind.Id, null, new Dictionary(), TaskSchedule.Continuous, true, now, 0), ct); } - - await _settings.SetAsync("tasks.seeded", "1", ct); } public async ValueTask> CreateAsync(TaskInput input, CancellationToken ct) diff --git a/backend/src/MoShiro.Backend.Gateway/FleetGateway.cs b/backend/src/MoShiro.Backend.Gateway/FleetGateway.cs index d8fae4d..55d4fb4 100644 --- a/backend/src/MoShiro.Backend.Gateway/FleetGateway.cs +++ b/backend/src/MoShiro.Backend.Gateway/FleetGateway.cs @@ -67,6 +67,9 @@ public sealed class FleetGateway : IAgentGateway public async ValueTask NotifyUpdateOfferAsync(NodeId node, string manifest, string signature, CancellationToken ct) => await Connections.NotifyAsync(node, MoShiroMethods.UpdateOffer, new UpdateOfferParams { Manifest = manifest, Signature = signature }, MoShiroJsonContext.Default.UpdateOfferParams, ct); + public async ValueTask NotifyScheduleAsync(NodeId node, string? restartAt, CancellationToken ct) => + await Connections.NotifyAsync(node, MoShiroMethods.Schedule, new ScheduleParams { RestartAt = restartAt }, MoShiroJsonContext.Default.ScheduleParams, ct); + public async ValueTask NotifyUpdateApplyAsync(NodeId node, CancellationToken ct) => await Connections.NotifyAsync(node, MoShiroMethods.UpdateApply, new UpdateOfferParams(), MoShiroJsonContext.Default.UpdateOfferParams, ct); } diff --git a/backend/tests/MoShiro.Backend.Tests/AgentEndToEndTests.cs b/backend/tests/MoShiro.Backend.Tests/AgentEndToEndTests.cs index 7f2e0ec..1cfbc88 100644 --- a/backend/tests/MoShiro.Backend.Tests/AgentEndToEndTests.cs +++ b/backend/tests/MoShiro.Backend.Tests/AgentEndToEndTests.cs @@ -44,7 +44,7 @@ public sealed class AgentEndToEndTests { // 默认模板的端口（7443/7444）不在这个 Agent 实际监听的端口里：任务被拒绝，原因是 policy_denied await using AgentHarness agent = await AgentHarness.StartAsync(_backend); - await BackendFixture.EventuallyAsync(async () => (await AgentViewAsync(agent))["rejected"]!.AsArray().Count == 2); + await BackendFixture.EventuallyAsync(async () => (await AgentViewAsync(agent))["rejected"]!.AsArray().Count == Defaults.TaskCount); JsonNode rejected = (await AgentViewAsync(agent))["rejected"]!; Assert.All(rejected.AsArray(), item => Assert.Equal("policy_denied", item!["reason"]!.GetValue())); diff --git a/backend/tests/MoShiro.Backend.Tests/AgentHarness.cs b/backend/tests/MoShiro.Backend.Tests/AgentHarness.cs index e83a6d8..e4318ab 100644 --- a/backend/tests/MoShiro.Backend.Tests/AgentHarness.cs +++ b/backend/tests/MoShiro.Backend.Tests/AgentHarness.cs @@ -88,16 +88,19 @@ public sealed class AgentHarness : IAsyncDisposable public async Task OverrideLineTasksAsync(int? rateHz = null) { JsonNode tasks = await Backend.GetJsonAsync(Backend.Admin, "/api/admin/tasks"); - foreach (JsonNode? task in tasks["items"]!.AsArray()) + // 先处理 connect（关掉），再改 udp_echo / tcp_echo 的速率：否则中途的带宽预算会多出 connect 那一份 + foreach (JsonNode? task in tasks["items"]!.AsArray().OrderBy(t => t!["kind"]!.GetValue() == "connect" ? 0 : 1)) { - if (task!["scope"]!.GetValue() != "line" || task["agent_id"] is not null || task["kind"]!.GetValue() is not ("udp_echo" or "tcp_echo")) + if (task!["scope"]!.GetValue() != "line" || task["agent_id"] is not null || task["kind"]!.GetValue() is not ("udp_echo" or "tcp_echo" or "connect")) { continue; } string kind = task["kind"]!.GetValue(); var parameters = new JsonObject { ["port"] = kind == "udp_echo" ? App.UdpPorts[0] : App.TcpPorts[0] }; - if (rateHz is { } hz) + // 默认的 connect 模板在这些用例里不需要（也占预算）：对这个 Agent 关掉它 + bool enabled = kind != "connect"; + if (rateHz is { } hz && kind != "connect") { parameters["rate_hz"] = hz; } @@ -109,6 +112,7 @@ public sealed class AgentHarness : IAsyncDisposable ["agent_id"] = Id, ["base_task_id"] = task["id"]!.GetValue(), ["params"] = parameters, + ["enabled"] = enabled, })); Assert.True(created.StatusCode == System.Net.HttpStatusCode.Created, await created.Content.ReadAsStringAsync()); } diff --git a/backend/tests/MoShiro.Backend.Tests/BackendFixture.cs b/backend/tests/MoShiro.Backend.Tests/BackendFixture.cs index 47120ab..95f4397 100644 --- a/backend/tests/MoShiro.Backend.Tests/BackendFixture.cs +++ b/backend/tests/MoShiro.Backend.Tests/BackendFixture.cs @@ -10,6 +10,12 @@ using MoShiro.TestSupport; namespace MoShiro.Backend.Tests; +/// 首次启动种入的默认线路任务模板数：udp_echo、tcp_echo、connect。 +internal static class Defaults +{ + public const int TaskCount = 3; +} + /// /// 在真实的 Kestrel 端口上启动整个后端（真实的 SQLite 与段文件，只是用临时目录和较快的轮询）， /// 并提供管理员客户端（Bearer 令牌）和假 Agent。 diff --git a/backend/tests/MoShiro.Backend.Tests/EnrollmentAndTaskSetTests.cs b/backend/tests/MoShiro.Backend.Tests/EnrollmentAndTaskSetTests.cs index b97414e..b6c2213 100644 --- a/backend/tests/MoShiro.Backend.Tests/EnrollmentAndTaskSetTests.cs +++ b/backend/tests/MoShiro.Backend.Tests/EnrollmentAndTaskSetTests.cs @@ -24,8 +24,8 @@ public sealed class EnrollmentAndTaskSetTests Assert.Equal("active", agent.EnrollmentStatus); Assert.Equal("active", agent.AuthStatus); - TaskSet set = await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); - Assert.Equal(["tcp_echo", "udp_echo"], set.Tasks.Select(t => t.Kind).Order().ToArray()); + TaskSet set = await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); + Assert.Equal(["connect", "tcp_echo", "udp_echo"], set.Tasks.Select(t => t.Kind).Order().ToArray()); Assert.All(set.Tasks, t => { Assert.Equal("line", t.Scope); @@ -56,7 +56,7 @@ public sealed class EnrollmentAndTaskSetTests public async Task Reconnecting_AlwaysGetsTheCurrentSnapshotAgain() { AgentSimulator first = await _backend.ConnectAgentAsync(); - TaskSet before = await first.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + TaskSet before = await first.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); await first.DisposeAsync(); // 同一个身份重新连接：新连接的服务端状态为空，所以会重新推送当前版本 @@ -157,7 +157,7 @@ public sealed class EnrollmentAndTaskSetTests using HttpResponseMessage approve = await _backend.Admin.PostAsync($"/api/admin/agents/{agent.NodeId.Value}/approve", null); Assert.Equal(HttpStatusCode.OK, approve.StatusCode); - TaskSet set = await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + TaskSet set = await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); Assert.NotNull(set); } @@ -165,7 +165,7 @@ public sealed class EnrollmentAndTaskSetTests public async Task DisablingAConnectedAgent_RevokesItAndRefusesFutureConnections() { await using AgentSimulator agent = await _backend.ConnectAgentAsync(); - await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); using HttpResponseMessage disable = await _backend.Admin.PostAsync($"/api/admin/agents/{agent.NodeId.Value}/disable", null); Assert.Equal(HttpStatusCode.OK, disable.StatusCode); @@ -239,7 +239,7 @@ public sealed class EnrollmentAndTaskSetTests public async Task TaskOverride_ChangesTheDesiredStateAndBumpsTheVersion() { await using AgentSimulator agent = await _backend.ConnectAgentAsync(); - TaskSet initial = await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + TaskSet initial = await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); TaskSpec udp = initial.Tasks.Single(t => t.Kind == "udp_echo"); using HttpResponseMessage created = await _backend.Admin.PostAsync("/api/admin/tasks", BackendFixture.Json(new JsonObject @@ -254,7 +254,7 @@ public sealed class EnrollmentAndTaskSetTests Assert.StartsWith("/api/admin/tasks/", created.Headers.Location!.ToString()); TaskSet updated = await agent.WaitForTaskSetAsync(s => s.Version > initial.Version && s.Tasks.Any(t => t.Kind == "udp_echo" && t.Params["rate_hz"].GetInt64() == 20)); - Assert.Equal(2, updated.Tasks.Count); + Assert.Equal(Defaults.TaskCount, updated.Tasks.Count); Assert.Equal(7443, updated.Tasks.Single(t => t.Kind == "udp_echo").Params["port"].GetInt64()); // 停用覆盖 → 该 Agent 不再运行这条任务（模板对其他 Agent 不受影响） @@ -262,7 +262,7 @@ public sealed class EnrollmentAndTaskSetTests using HttpResponseMessage disable = await _backend.Admin.PatchAsync("/api/admin/tasks/" + overrideId, BackendFixture.Json(new JsonObject { ["enabled"] = false }, "application/merge-patch+json")); Assert.Equal(HttpStatusCode.OK, disable.StatusCode); TaskSet disabled = await agent.WaitForTaskSetAsync(s => s.Version > updated.Version && s.Tasks.All(t => t.Kind != "udp_echo")); - Assert.Single(disabled.Tasks); + Assert.Equal(Defaults.TaskCount - 1, disabled.Tasks.Count); } [Fact] @@ -314,7 +314,7 @@ public sealed class EnrollmentAndTaskSetTests public async Task TaskSet_RejectedItemsReportedByTheAgent_AreVisibleToTheAdmin() { await using AgentSimulator agent = await _backend.ConnectAgentAsync(); - await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); agent.RejectNextState = [new Fleet.Rpc.RejectedItem { Id = "tsk_x", Reason = "policy_denied" }]; using HttpResponseMessage created = await _backend.Admin.PostAsync("/api/admin/tasks", BackendFixture.Json(new JsonObject @@ -330,7 +330,7 @@ public sealed class EnrollmentAndTaskSetTests public async Task ProbeLifecycle_CreateRotateRevokeDelete_IsReflectedInEveryAgentsPeerList() { await using AgentSimulator agent = await _backend.ConnectAgentAsync(); - await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); using HttpResponseMessage created = await _backend.Admin.PostAsync("/api/admin/probes", BackendFixture.Json(new JsonObject { ["name"] = _backend.UniqueName("home") })); Assert.Equal(HttpStatusCode.Created, created.StatusCode); diff --git a/backend/tests/MoShiro.Backend.Tests/OnDemandTestsTests.cs b/backend/tests/MoShiro.Backend.Tests/OnDemandTestsTests.cs index 66f200a..fc9fde4 100644 --- a/backend/tests/MoShiro.Backend.Tests/OnDemandTestsTests.cs +++ b/backend/tests/MoShiro.Backend.Tests/OnDemandTestsTests.cs @@ -25,7 +25,7 @@ public sealed class OnDemandTestsTests private async Task AgentWithProbeAsync() { AgentSimulator agent = await _backend.ConnectAgentAsync(); - await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); await agent.SendEventAsync(AgentEventKinds.ProbeConnected, _backend.DefaultProbeId); await BackendFixture.EventuallyAsync(async () => (await _backend.GetJsonAsync(_backend.Admin, "/api/admin/agents/" + agent.NodeId.Value))["probe_connected"]!.GetValue()); return agent; @@ -37,7 +37,7 @@ public sealed class OnDemandTestsTests public async Task Start_RequiresAProbeSession_APlanThatExists_AndValidParams() { await using AgentSimulator agent = await _backend.ConnectAgentAsync(); - await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); using HttpResponseMessage offline = await StartAsync(agent, Plan("hy2_ladder")); Assert.Equal(HttpStatusCode.Conflict, offline.StatusCode); diff --git a/backend/tests/MoShiro.Backend.Tests/OperationsAndSecurityTests.cs b/backend/tests/MoShiro.Backend.Tests/OperationsAndSecurityTests.cs index c7b6215..465b79a 100644 --- a/backend/tests/MoShiro.Backend.Tests/OperationsAndSecurityTests.cs +++ b/backend/tests/MoShiro.Backend.Tests/OperationsAndSecurityTests.cs @@ -139,7 +139,7 @@ public sealed class OperationsAndSecurityTests public async Task OversizedFrames_CloseTheConnection_WithoutHarmingTheServer() { await using AgentSimulator agent = await _backend.ConnectAgentAsync(); - await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); try { await agent.SendBinaryAsync(new byte[17 * 1024 * 1024]); @@ -153,14 +153,14 @@ public sealed class OperationsAndSecurityTests // 服务端仍然正常 await using AgentSimulator another = await _backend.ConnectAgentAsync(); - Assert.NotNull(await another.WaitForTaskSetAsync(s => s.Tasks.Count == 2)); + Assert.NotNull(await another.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount)); } [Fact] public async Task MalformedAndUnknownMessages_AreIgnoredOrRefused_AndTheConnectionStaysUsable() { await using AgentSimulator agent = await _backend.ConnectAgentAsync(); - await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); await agent.SendTextAsync("{not json"); await agent.SendTextAsync("[]"); @@ -187,7 +187,7 @@ public sealed class OperationsAndSecurityTests Assert.Empty(agent.TaskSets); await agent.HelloAsync(); - Assert.NotNull(await agent.WaitForTaskSetAsync(s => s.Tasks.Count == 2)); + Assert.NotNull(await agent.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount)); } [Fact] @@ -205,9 +205,9 @@ public sealed class OperationsAndSecurityTests public async Task ReconnectingTheSameNode_ReplacesTheOldConnection() { AgentSimulator first = await _backend.ConnectAgentAsync(); - await first.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await first.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); await using AgentSimulator second = await _backend.ReconnectAsync(first); - await second.WaitForTaskSetAsync(s => s.Tasks.Count == 2); + await second.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount); // 旧连接被替换掉，但节点仍然在线（在线状态属于新连接） await first.WaitUntilAsync(() => first.CloseReason is not null); diff --git a/backend/tests/MoShiro.Backend.Tests/PersistenceTests.cs b/backend/tests/MoShiro.Backend.Tests/PersistenceTests.cs index 9c60385..87be2b7 100644 --- a/backend/tests/MoShiro.Backend.Tests/PersistenceTests.cs +++ b/backend/tests/MoShiro.Backend.Tests/PersistenceTests.cs @@ -57,7 +57,7 @@ public sealed class PersistenceTests : IAsyncLifetime // 同一个身份可以重新连接，并拿回当前版本的期望状态 await using AgentSimulator again = await _backend.ReconnectAsync(agent); - Assert.NotNull(await again.WaitForTaskSetAsync(s => s.Tasks.Count == 2)); + Assert.NotNull(await again.WaitForTaskSetAsync(s => s.Tasks.Count == Defaults.TaskCount)); // 已经落盘的块在重启后仍然可以按 ID 下载 JsonNode chunks = await _backend.GetJsonAsync(admin, $"/api/admin/raw/chunks?agent_id={id}&limit=1"); diff --git a/backend/tests/MoShiro.Backend.Tests/ScheduleTests.cs b/backend/tests/MoShiro.Backend.Tests/ScheduleTests.cs new file mode 100644 index 0000000..fe315bb --- /dev/null +++ b/backend/tests/MoShiro.Backend.Tests/ScheduleTests.cs @@ -0,0 +1,80 @@ +using System.Net; +using System.Text.Json.Nodes; +using MoShiro.Backend.Application; +using MoShiro.Protocol; +using MoShiro.TestSupport; + +namespace MoShiro.Backend.Tests; + +/// 定时重启：默认开、北京时间 04:30；节点连上时补发，管理端改了设置马上群发，关掉就发空。 +[Collection("probe-e2e")] +public sealed class ScheduleTests +{ + private readonly BackendFixture _backend; + + public ScheduleTests(BackendFixture backend) + { + _backend = backend; + } + + private async Task PatchAsync(JsonObject settings) + { + using HttpResponseMessage response = await _backend.Admin.PatchAsync("/api/admin/settings", BackendFixture.Json(settings, "application/merge-patch+json")); + Assert.Equal(HttpStatusCode.OK, response.StatusCode); + } + + private static Task LastIs(AgentSimulator agent, string? restartAt) => + Task.FromResult(agent.Schedules.ToArray() is { Length: > 0 } all && all[^1].RestartAt == restartAt); + + [Fact] + public void Format_WritesTheOffsetTheNodeParses() + { + Assert.Equal("04:30+08:00", ScheduleService.Format(4, 30, 480)); + Assert.Equal("00:00+00:00", ScheduleService.Format(0, 0, 0)); + Assert.Equal("23:05-05:30", ScheduleService.Format(23, 5, -330)); + } + + [Fact] + public async Task ANewNode_GetsTheDefaultSchedule_AndLaterChangesAreBroadcast() + { + await using AgentSimulator agent = await _backend.ConnectAgentAsync(); + await BackendFixture.EventuallyAsync(() => LastIs(agent, "04:30+08:00"), because: "defaults: on, 04:30 Beijing time, sent right after hello"); + + try + { + await PatchAsync(new JsonObject { ["fleet.restart_hour"] = 3, ["fleet.restart_minute"] = 5, ["fleet.restart_utc_offset_minutes"] = 0 }); + await BackendFixture.EventuallyAsync(() => LastIs(agent, "03:05+00:00"), because: "a changed time is pushed to online nodes"); + + await PatchAsync(new JsonObject { ["fleet.restart_enabled"] = false }); + await BackendFixture.EventuallyAsync(() => LastIs(agent, null), because: "switching it off sends an empty schedule"); + + // 关着的时候新连上的节点也拿到「关」 + await using AgentSimulator late = await _backend.ConnectAgentAsync(); + await BackendFixture.EventuallyAsync(() => LastIs(late, null), because: "a node connecting while it is off is told so"); + } + finally + { + await PatchAsync(new JsonObject { ["fleet.restart_enabled"] = true, ["fleet.restart_hour"] = 4, ["fleet.restart_minute"] = 30, ["fleet.restart_utc_offset_minutes"] = 480 }); + } + } + + [Fact] + public async Task BadValues_AreRefused() + { + using HttpResponseMessage response = await _backend.Admin.PatchAsync("/api/admin/settings", + BackendFixture.Json(new JsonObject { ["fleet.restart_hour"] = 24, ["fleet.restart_minute"] = 60, ["fleet.restart_utc_offset_minutes"] = 900 }, "application/merge-patch+json")); + Assert.Equal(HttpStatusCode.BadRequest, response.StatusCode); + } + + [Fact] + public async Task DefaultLineTemplates_IncludeTheConnectTimings() + { + using HttpResponseMessage response = await _backend.Admin.GetAsync("/api/admin/tasks"); + Assert.Equal(HttpStatusCode.OK, response.StatusCode); + JsonArray items = JsonNode.Parse(await response.Content.ReadAsStringAsync())!["items"]!.AsArray(); + string[] kinds = [.. items.Select(i => i!["kind"]!.GetValue())]; + Assert.Contains("connect", kinds); + Assert.Contains("udp_echo", kinds); + Assert.Contains("tcp_echo", kinds); + } +} diff --git a/backend/tests/MoShiro.TestSupport/AgentSimulator.cs b/backend/tests/MoShiro.TestSupport/AgentSimulator.cs index 15b0d30..13ffdf0 100644 --- a/backend/tests/MoShiro.TestSupport/AgentSimulator.cs +++ b/backend/tests/MoShiro.TestSupport/AgentSimulator.cs @@ -51,6 +51,9 @@ public sealed class AgentSimulator : IAsyncDisposable /// 面板发来的发布清单（moshiro.update.offer）和更新请求（moshiro.update.apply）。 public ConcurrentQueue UpdateOffers { get; } = new(); + /// 面板下发的定时重启设置（moshiro.schedule），按到达顺序。 + public ConcurrentQueue Schedules { get; } = new(); + private int _updateApplies; public int UpdateApplies => Volatile.Read(ref _updateApplies); @@ -406,6 +409,9 @@ public sealed class AgentSimulator : IAsyncDisposable case MoShiroMethods.UpdateOffer: UpdateOffers.Enqueue(message.Params!.Value.Deserialize(MoShiroJsonContext.Default.UpdateOfferParams)!); break; + case MoShiroMethods.Schedule: + Schedules.Enqueue(message.Params!.Value.Deserialize(MoShiroJsonContext.Default.ScheduleParams)!); + break; case MoShiroMethods.UpdateApply: Interlocked.Increment(ref _updateApplies); break; diff --git a/docs/api.md b/docs/api.md index 8449434..d21a034 100644 --- a/docs/api.md +++ b/docs/api.md @@ -142,7 +142,7 @@ | GET | `/api/admin/events?agent_id=&kind=&from=&to=` | 事件 | | GET | `/api/admin/analysis/recompute` | 指标重算进度 | | GET/POST/DELETE | `/api/admin/api-tokens` | 作用域为 `read`、`admin` 或 `release`（只能登记发布清单，给 CI 用，什么都读不了）；明文只显示一次 | -| GET/PATCH | `/api/admin/settings` | 站点公开或私有、是否对访客显示构建信息（`site.show_build`）、保留期等 | +| GET/PATCH | `/api/admin/settings` | 站点公开或私有、是否对访客显示构建信息（`site.show_build`）、保留期、节点定时重启（`fleet.restart_enabled`、`fleet.restart_hour`、`fleet.restart_minute`、`fleet.restart_utc_offset_minutes`，默认开、04:30、UTC+8；改完立即通知在线节点）等 | | GET | `/api/admin/build-info` | 后端和每个 Agent、Probe 的版本、tag、提交、构建时间、系统，以及节点宿主机更新器报告的更新状态（管理端「系统 → 版本」） | | GET/POST | `/api/admin/releases` | CI 签名的发布清单：POST 登记（`{manifest, signature}`，面板不验签、原样转给在线的节点和之后连上的节点），GET 看每个组件最新的一份 | | GET | `/api/admin/links` | 每一对（Probe，Agent）的线路开关：`linked`（最终结果）、`override`（手动设置，`null` 跟随自动连接）、`connected`，以及 `auto_link` | diff --git a/docs/deployment.md b/docs/deployment.md index 4c4489b..33ae25a 100644 --- a/docs/deployment.md +++ b/docs/deployment.md @@ -238,6 +238,15 @@ Backend 用根目录的 `compose.yaml`（见上文「Backend 容器」），数 已在 Fedora CoreOS 的 SELinux Enforcing 环境、Podman 6.0.2 rootless 下用同一份 Compose 实测：Agent 注册、宿主指标、Probe 登录和 TLS 连接正常，IPv4/IPv6 ICMP 目标任务收到有效回包，容器保持 `cap_drop: ALL`，未添加 NET_RAW。Linux ICMP 优先使用 datagram ping socket，不依赖外部 ping 程序；需要 `net.ipv4.ping_group_range` 包含容器映射后的 GID（测试环境为 `0 2147483647`）。宿主禁止普通用户 ping 时，需管理员配置该范围；例如 rootful Docker 的用户 GID 1654 可设置 `sysctl -w "net.ipv4.ping_group_range=1654 1654"`。rootless 应使用映射后的宿主 GID，不能直接套用容器 GID。单加 `cap_add: NET_RAW` 不保证非 root 用户获得有效 capability。有原始套接字权限的部署仍保留原有 Ping 回退路径。其他操作系统继续使用 System.Net.Ping。ICMP 目标任务当前由 Agent 执行，Probe 没有独立目标任务入口。 +### 定时重启 + +Agent 和 Probe 每天固定时刻有序退出一次，由服务管理器拉起（所有 systemd 单元和 OpenWrt 的 procd 脚本都是「总是重启」，正常退出也会被拉起）。节点只是自己退出，没有任何执行命令的能力；重启期间测量会中断几十秒。 + +- **在管理端控制**：系统 → 设置 →「定时重启」，开关、时刻和时区（默认**开**，北京时间 04:30）。保存后马上通知所有在线节点，新连上的节点会补发。 +- **节点本地覆盖**：节点本地写了就以本地为准（面板的设置对它不生效）。`agent.json` 或 `probe.json` 里加 `"restart_at": "04:30+08:00"`，或者设置环境变量 `MOSHIRO_RESTART_AT`（优先）。写法 `HH:mm`（按 UTC）、`HH:mmZ` 或 `HH:mm±hh:mm`；写错了启动时就会报错。从面板重新下载 `probe.json` 会丢掉手写的 `restart_at`，所以 Probe 的容器推荐用环境变量。 +- 启动后不满 10 分钟不会触发，所以重启之后不会立刻又重启一次。日志里有 `restart scheduled at=… source=panel|local` 和 `stopping result=scheduled_restart`。 +- Windows 上没有自带的服务单元，需要服务管理器配置为退出后重启。 + ### Agent 的系统指标挂载 Agent 的 Compose 模板已经包含以下只读挂载和环境变量，systemd 示例也采用相同配置： diff --git a/docs/measurement.md b/docs/measurement.md index 0a56449..a581eb2 100644 --- a/docs/measurement.md +++ b/docs/measurement.md @@ -81,7 +81,7 @@ public static readonly KindSpec TcpEcho = new( | `hold` | tls + hold | 每 30 s 一次心跳 | reality | | `icmp_echo` | icmp + fixed + fixed + echo | 1 Hz × 64 B | 目标任务 | -默认的线路任务模板：`udp_echo`、`tcp_echo`（持续运行）。其余 Kind 可以按需开启，或者以 Plan 形式运行。 +默认的线路任务模板：`udp_echo`、`tcp_echo`、`connect`（持续运行；`connect` 每分钟 6 次，记录 TCP 建连和 TLS 握手耗时，约 0.4 kbps）。其余 Kind 可以按需开启，或者以 Plan 形式运行。每一批默认任务只种入一次，在设置里用标记记住，所以删掉的默认模板不会被重新建出来；后来新增的默认任务（`connect`）在已有部署升级后会补建一次。 ### 3.4 回显的四个时间戳 diff --git a/docs/protocol.md b/docs/protocol.md index 32cf03f..c982a95 100644 --- a/docs/protocol.md +++ b/docs/protocol.md @@ -13,6 +13,7 @@ MoShiro 的两条控制链路都建立在 Fleet 之上： | --- | --- | --- | | 下发任务 | 期望状态 `fleet.state.sync` | 文档类型 `TaskSet`，包括 `tasks[]`、`peers`（允许连这个 Agent 的 Probe 公钥，已按线路开关过滤） | | 下发 Agent 列表给 Probe | 期望状态 `fleet.state.sync`（Probe 直连面板） | 文档类型 `probe_agents`：`version`、`agents[]`（`agent_id`、`host`、`tcp_port`、`udp_ports`、`spki`），已按线路开关过滤；Agent 集合或线路开关变化时重新推送 | +| 定时重启 | 通知 `moshiro.schedule`（B→A、B→P） | `{restart_at}`，形如 `04:30+08:00`，空表示关闭；连上时补发、管理端改了设置就群发。节点本地写了 `restart_at` 或 `MOSHIRO_RESTART_AT` 时以本地为准；节点只是到点自己有序退出，由服务管理器拉起 | | 更新 | 通知 `moshiro.update.offer` / `moshiro.update.apply`（B→A、B→P），心跳附加字段 `update` | `offer` 带 CI 签名的发布清单原文和签名，节点只写进数据目录；`apply` 请求更新；`update` 是宿主机更新器的状态（`state`、`current`、`available`、`error`、`at`），见 [deployment.md](deployment.md) | | 上报原始块 | 发件箱，二进制通道 **16** | `RawChunk`（自描述，见 [measurement.md](measurement.md) 第 4 节），`msg_id = chunk_id` | | Probe 交付列块 | 点对点会话的发件箱，二进制通道 **17** | 列块：`u32 stream_id \| u32 first_seq \| u32 count \| 列数据(Brotli)` | diff --git a/web/admin/src/pages/System.tsx b/web/admin/src/pages/System.tsx index f47de95..60a088a 100644 --- a/web/admin/src/pages/System.tsx +++ b/web/admin/src/pages/System.tsx @@ -10,70 +10,166 @@ import { Chip, Confirm, CopyButton, Dialog, Empty, ErrorNotice, Field, Segmented type Tab = 'settings' | 'versions' | 'tokens' | 'users' | 'events' | 'raw' | 'analysis' +const RESTART_KEYS = ['fleet.restart_enabled', 'fleet.restart_hour', 'fleet.restart_minute', 'fleet.restart_utc_offset_minutes'] + +/** 常用时区（固定偏移，不含夏令时）：偏移是东经分钟数。 */ +const TIME_ZONES: { offset: number; label: string }[] = [ + { offset: 480, label: '北京时间（UTC+8）' }, + { offset: 0, label: 'UTC' }, + { offset: 540, label: '东京、首尔（UTC+9）' }, + { offset: 60, label: '中欧（UTC+1）' }, + { offset: 180, label: '莫斯科（UTC+3）' }, + { offset: -300, label: '美东（UTC-5）' }, + { offset: -480, label: '美西（UTC-8）' }, +] + +const pad2 = (n: number) => String(n).padStart(2, '0') +const offsetLabel = (offset: number) => `UTC${offset < 0 ? '-' : '+'}${Math.floor(Math.abs(offset) / 60)}${Math.abs(offset) % 60 ? `:${pad2(Math.abs(offset) % 60)}` : ''}` + +/** 定时重启：节点每天这个时刻有序退出一次，由系统服务拉起。设置保存后马上通知所有在线节点。 */ +function RestartCard({ items, onSaved }: { items: SettingDto[]; onSaved: () => void }) { + const { busy, run } = useAction() + const value = (key: string, fallback: string) => items.find((s) => s.key === key)?.value ?? fallback + const saved = { + enabled: value('fleet.restart_enabled', 'true') === 'true', + hour: Number(value('fleet.restart_hour', '4')), + minute: Number(value('fleet.restart_minute', '30')), + offset: Number(value('fleet.restart_utc_offset_minutes', '480')), + } + const [draft, setDraft] = useState(saved) + useEffect(() => setDraft(saved), [saved.enabled, saved.hour, saved.minute, saved.offset]) // eslint-disable-line react-hooks/exhaustive-deps + const dirty = draft.enabled !== saved.enabled || draft.hour !== saved.hour || draft.minute !== saved.minute || draft.offset !== saved.offset + const zones = TIME_ZONES.some((z) => z.offset === draft.offset) ? TIME_ZONES : [...TIME_ZONES, { offset: draft.offset, label: offsetLabel(draft.offset) }] + const zoneName = zones.find((z) => z.offset === draft.offset)?.label ?? offsetLabel(draft.offset) + + return ( +
+
定时重启（Agent 和 Probe）
+
+ + + + : + + + + +
+

+ {draft.enabled ? `每天 ${pad2(draft.hour)}:${pad2(draft.minute)}（${zoneName}）` : '已关闭'} + ：节点自己有序退出一次，由系统服务（systemd、容器、procd）重新拉起，测量会中断几十秒。节点本地写了 restart_at 或环境变量 MOSHIRO_RESTART_AT 时以本地为准。 +

+
+ ) +} + function Settings() { const settings = useQuery(() => api.get>('/api/admin/settings'), []) const { busy, run } = useAction() const [draft, setDraft] = useState>({}) useEffect(() => setDraft({}), [settings.data]) - const items = settings.data?.items ?? [] + const items = (settings.data?.items ?? []).filter((s) => !RESTART_KEYS.includes(s.key)) const dirty = Object.keys(draft).length > 0 return ( -
- - - - - - - - - - - {items.map((s) => { - const value = draft[s.key] ?? s.value - return ( - - - - - - ) - })} - -
设置值说明
{s.key} - {s.type === 'bool' ? ( - - ) : ( - setDraft({ ...draft, [s.key]: e.target.value })} /> - )} - - {translateSetting(s.key, s.description)} - {s.value !== s.default ? （默认 {s.default}） : null} -
-
- - 「site.public」打开之后，展示页对匿名访客公开。 + <> + +
+ + + + + + + + + + + {items.map((s) => { + const value = draft[s.key] ?? s.value + return ( + + + + + + ) + })} + +
设置值说明
{s.key} + {s.type === 'bool' ? ( + + ) : ( + setDraft({ ...draft, [s.key]: e.target.value })} /> + )} + + {translateSetting(s.key, s.description)} + {s.value !== s.default ? （默认 {s.default}） : null} +
+
+ + 「site.public」打开之后，展示页对匿名访客公开。 +
-
+ ) } diff --git a/web/display/src/pages/Line.tsx b/web/display/src/pages/Line.tsx index 136f28b..c57fa6c 100644 --- a/web/display/src/pages/Line.tsx +++ b/web/display/src/pages/Line.tsx @@ -30,7 +30,7 @@ interface HostSeries { } /** 指标卡片：好坏的分界（越低越好）和显示格式。 */ -const TILES: { label: string; key: string; fmt: (v: number | null | undefined) => string; good?: number; poor?: number; unit?: string }[] = [ +const TILES: { label: string; key: string; fmt: (v: number | null | undefined) => string; good?: number; poor?: number; unit?: string; /** 这个线路没有数据时不显示（建连任务不是每条线路都开着）。 */ optional?: boolean }[] = [ { label: 'UDP 延迟 p50', key: 'rtt.p50', fmt: fmtMs }, { label: 'UDP 延迟 p95', key: 'rtt.p95', fmt: fmtMs }, { label: 'UDP 丢包', key: 'loss.total', fmt: fmtPct, good: 0.5, poor: 3 }, @@ -43,6 +43,9 @@ const TILES: { label: string; key: string; fmt: (v: number | null | undefined) = { label: 'TCP RTT p99', key: 'tcp.rtt.p99', fmt: fmtMs }, { label: 'TCP RTT 最大', key: 'tcp.rtt.max', fmt: fmtMs }, { label: 'TCP 重传率', key: 'tcp.retrans', fmt: fmtPct, good: 0.5, poor: 3 }, + { label: 'TCP 建连 p95', key: 'connect.tcp_ms.p95', fmt: fmtMs, optional: true }, + { label: 'TLS 握手 p95', key: 'connect.tls_ms.p95', fmt: fmtMs, optional: true }, + { label: '建连成功率', key: 'connect.success', fmt: fmtPct, optional: true }, ] interface ChartDef { @@ -52,7 +55,7 @@ interface ChartDef { zero?: boolean fill?: boolean series: { key: string; label: string; color: string; dash?: number[] }[] - group: 'udp' | 'tcp' | 'score' + group: 'udp' | 'tcp' | 'connect' | 'score' /** 占满一行。 */ wide?: boolean } @@ -60,6 +63,7 @@ interface ChartDef { const GROUPS = [ { id: 'udp', title: 'UDP（往返探测）' }, { id: 'tcp', title: 'TCP' }, + { id: 'connect', title: '建连（TCP 握手、TLS 握手）' }, { id: 'score', title: '评分' }, ] as const @@ -69,6 +73,8 @@ const CHARTS: ChartDef[] = [ { id: 'jitter', title: '抖动（p95）', group: 'udp', unit: 'ms', series: [{ key: 'jitter.up.p95', label: '上行', color: '--chart-2' }, { key: 'jitter.down.p95', label: '下行', color: '--chart-1' }] }, { id: 'tcp', title: '往返时延（RTT）', group: 'tcp', unit: 'ms', series: [{ key: 'tcp.rtt.p50', label: 'p50', color: '--chart-1' }, { key: 'tcp.rtt.p95', label: 'p95', color: '--chart-2' }, { key: 'tcp.rtt.max', label: '最大', color: '--chart-3' }] }, { id: 'retrans', title: '重传率', group: 'tcp', unit: '%', fill: true, series: [{ key: 'tcp.retrans', label: '重传率', color: '--chart-3' }] }, + { id: 'connect-time', title: '建连耗时（p95）', group: 'connect', unit: 'ms', series: [{ key: 'connect.tcp_ms.p95', label: 'TCP 建连', color: '--chart-1' }, { key: 'connect.tls_ms.p95', label: 'TLS 握手', color: '--chart-2' }] }, + { id: 'connect-ok', title: '建连成功率', group: 'connect', unit: '%', zero: false, series: [{ key: 'connect.success', label: '成功率', color: '--chart-3' }] }, { id: 'score', title: '综合评分', group: 'score', wide: true, unit: '', series: [{ key: 'score.reality', label: 'REALITY', color: '--chart-1' }, { key: 'score.hy2', label: 'Hy2', color: '--chart-2' }, { key: 'score.mc', label: 'MC', color: '--chart-4' }, { key: 'score.base', label: '综合', color: '--chart-5', dash: [4, 3] }] }, ] @@ -347,7 +353,7 @@ export function Line() { 评分
- {TILES.map((t) => ( + {TILES.filter((t) => !t.optional || m[t.key] != null).map((t) => ( ))} diff --git a/web/display/src/periods.ts b/web/display/src/periods.ts index cd503c2..1090d72 100644 --- a/web/display/src/periods.ts +++ b/web/display/src/periods.ts @@ -173,7 +173,7 @@ export function usePeriod(lineId: string, ready: boolean, sinceMs: number | null /** 这一段的序列；今天的「最近一段」会定时刷新。 */ export function usePeriodSeries(lineId: string, ready: boolean, period: Period) { return useQuery( - () => (ready ? api.get(`/api/public/lines/${encodeURIComponent(lineId)}/series`, { ...period.query, preset: 'udp,tcp,scores' }) : Promise.resolve(undefined)), + () => (ready ? api.get(`/api/public/lines/${encodeURIComponent(lineId)}/series`, { ...period.query, preset: 'udp,tcp,connect,scores' }) : Promise.resolve(undefined)), [lineId, JSON.stringify(period.query), ready], period.isToday ? (period.win === 'day' ? 60_000 : 10_000) : undefined, )