ModernUO/Projects/UOContent/Misc/CrowdSec/CrowdSecReporter.cs
Kamron Batman c39454137e
feat(network): pluggable connection filters; file blocklist + contribute-first CrowdSec (#2542)
Reshapes IP banning around one idea: **core owns the question, content owns every answer.**

Core gains a single accept-path seam — `IConnectionFilter` — and loses everything that used to implement one. The firewall moves to UOContent, a new file-backed blocklist joins it there, and CrowdSec is repositioned from an in-app enforcer to a contribute-first reporter.

## The seam

```csharp
public interface IConnectionFilter
{
    string Name { get; }
    void Configure();
    void Start(CancellationToken token);
    void Stop();
    bool ShouldDeny(IPAddress address);
}
```

The accept path went from hardcoded branches to one question:

```csharp
else if (ConnectionFilters.ShouldDeny(remoteIP, out var deniedBy))
{
    logger.Debug("{Address} denied by connection filter '{Filter}'", remoteIP, deniedBy);
}
```

Filters register during the Configure sweep. The registry is a plain array walked by an indexed loop — no enumerator, no closure, no allocation — and the first denial short-circuits. An interface dispatch is noise next to the `accept()` syscall, so pluggability costs nothing measurable on the path that has to survive a DDoS.

Whatever a hit implies — persisting, promoting to an OS bouncer, contributing to the ban channel — is the filter's business, not the accept path's.

A filter that throws is **unregistered and the connection fails open**. A filter that faults once faults for every subsequent connection, so leaving it registered means an exception and a log line per accept — exactly the amplification an attacker wants — and a broken filter must not be able to deny everyone either.

This deliberately does **not** reuse `EventSink.InvokeSocketConnect`: that fires later and allocates a `SocketConnectEventArgs` per connection, which is what the accept path avoids for rejected traffic.

## What ships behind it

**`firewall`** (UOContent) — the existing admin-curated set. Collapsed from `Firewall` + `AdminFirewall` + a threaded enforcer into one single-threaded store with **zero concurrency primitives**: the accept path, admin gump, TTL expiry and boot load all run on the game loop. Persists to `Configuration/firewall.json` with automatic migration from the legacy `firewall.cfg`. No behavior change for operators — same namespace, same gump, same commands.

**`blocklist`** (UOContent) — new. Holds a millions-strong list in-app and **demand-pages** hits up to CrowdSec, which promotes them to the OS firewall.

The motivation is concrete: CrowdSec's Windows bouncer cannot load the ~3.9M IPs that 91 community feeds produce, but it handles ~100k fine. So the millions live in-process behind a binary search, and only addresses that *actually connect* get promoted. A `PromotedGuard` suppresses re-reporting an address until the bouncer picks it up.

The list is parsed straight from UTF-8 file bytes with no per-line string allocation, off the game loop, and published as an immutable snapshot swapped through a single `volatile` reference. Reloads yield to world saves.

**`tools/Export-IpBlocklist.ps1`** — the producer. Requires PowerShell 7 and runs on Windows, Linux and macOS; Windows PowerShell 5.1 is refused up front via `#requires`. Merges a thin, non-overlapping feed set into one de-duplicated, bogon-filtered file. Parsing runs in a compiled `Add-Type` hot loop (~1s for ~4M lines instead of minutes). Written to a `.tmp` sibling and swapped with `File.Replace`, so the shard never reads a half-written list, and a total feed outage refuses to overwrite a good list with an empty one. Re-running is idempotent — it exits without downloading anything while the list on disk is younger than `-MinInterval` (default 2h, the anchor feed's own refresh period), so a misconfigured scheduler can't hammer upstream.

## CrowdSec: contribute-first

`IBanReporter` + `BanChannel` fan locally-decided bans out to external systems. `CrowdSecReporter` (UOContent) posts to LAPI `POST /v1/alerts` and retracts via `DELETE /v1/decisions`.

Reporting is **enqueue-only** on the accept path: a bounded, coalescing channel drained off-loop with bounded retry, counted drops on overflow, and a flush on shutdown. Under a DDoS the accept path never does synchronous or lock-contending per-IP work.

### Why not pull decisions from CrowdSec?

The original design streamed decisions into an in-app snapshot and enforced them at the accept gate. That's the wrong layer: by the time the shard sees the connection, the TCP handshake and socket setup are already paid for. `cs-firewall-bouncer` drops the same traffic **at the kernel**, and it's what CrowdSec is built to do. So the shard now contributes what it uniquely knows (rate-limit trips, blocklist hits from real connection attempts) and lets the OS enforce.

The one thing the OS can't do — hold millions of entries on Windows — is exactly what the in-app blocklist covers, and it feeds the same pipeline.

## Threading policy

`CLAUDE.md` rule #3 is rewritten as an explicit three-part policy, with rule #10 restated in tandem:

- Anything touching game state runs **only** on the main loop.
- Heavy work that *needs* game state must be **chunked** across ticks, never threaded.
- Heavy work that does *not* need game state (large-file parse, external I/O) **must** run off-loop **and must yield to world saves**.

Results come back via an immutable snapshot swapped through a single `volatile` reference, or `Core.LoopContext.Post` — never by letting the scheduler decide where heavy work runs. Both new subsystems follow it.

## Shared primitives

`SortedRangeIndex<T> where T : IBinaryInteger<T>` — coalesced disjoint interval arrays plus a binary search. The firewall, the blocklist, and (as of this PR) core's reserved-network tables all use it.

Coalescing is a correctness requirement, not an optimization: multi-feed lists nest CIDRs (`/24` containing a `/32`), and a search that inspects only the rightmost run whose minimum is ≤ the value is sound **only** over disjoint runs. That bug was caught in review and is covered by regression tests.

`IPAddressUtility` collects the allocation-free `IPAddress` ↔ `UInt128` conversions and CIDR parsing that were previously scattered or duplicated.

## Config

| File | Owner | Keys |
|---|---|---|
| `Configuration/bans.json` | core | `reportRateLimitTrips`, `autoBanDuration` |
| `Configuration/blocklist.json` | content | `file`, `reloadInterval`, `reportHits`, `banDuration`, `promoteSuppression` |
| `Configuration/crowdsec.json` | content | `lapiUrl`, `machineId`, `password`, `origin`, `manualBanDuration`, `flushInterval`, `maxQueue` |
| `Configuration/firewall.json` | content | persisted firewall entries (migrated from `firewall.cfg`) |

Everything is inert by default. CrowdSec self-disables without credentials; the blocklist self-disables until its file exists. A shard that changes nothing sees no behavior change.

## Notes for review

- **Core no longer references `Firewall` or `IFirewallEntry` anywhere.** `NetworkUtilities` used to build its reserved-network tables out of `CidrFirewallEntry`, which coupled core to the firewall for something unrelated to banning; those are now a `SortedRangeIndex<UInt128>`, same semantics and public API.
- **`BanChannel.Stop()` no longer persists the firewall** — a contribution coordinator has no business saving an enforcement store. That's the firewall filter's `Stop()`.
- **A dead `whitelisted` parameter was dropped** from the blocklist gate: it was hardcoded `false` at its only call site, and no whitelist concept exists in core.
- **The blocklist filter is an instance, not a static.** The static version forced its tests onto the sequential collection with a reset hook; they now run in parallel.
- `dev-docs/networking-packets.md` documents the seam for content authors, plus a known wart in the `IPAddress` ↔ `UInt128` normalization flagged for a follow-up PR.
- The generator was verified on Linux, macOS and Windows under a temporary CI matrix (since removed). It caught two portability bugs — a Windows-only path separator, and a culture-sensitive duration parse that read `2.5` as `25` on comma-decimal locales and *silently* turned a 2.5h cooldown into 25h — plus a third that made the script unparseable on Windows PowerShell 5.1. The source is ASCII-only for that last reason: `#requires` is only honored once a file parses, so non-ASCII in a BOM-less script produces parse errors instead of the version message.

## Tests

**1344 pass** (782 `Server.Tests`, 562 `UOContent.Tests`). New coverage: filter registry (registration, short-circuit, fault-disable), blocklist parsing/CIDR/coalescing, snapshot reload markers, promote-guard TTL, ban-channel fan-out, CrowdSec alert building/dedup/flush-on-stop, and the generator's output-format contract pinned against the reader.
2026-07-25 11:59:37 -07:00

399 lines
15 KiB
C#

/*************************************************************************
* ModernUO *
* Copyright 2019-2026 - ModernUO Development Team *
* Email: hi@modernuo.com *
* File: CrowdSecReporter.cs *
* *
* This program is free software: you can redistribute it and/or modify *
* it under the terms of the GNU General Public License as published by *
* the Free Software Foundation, either version 3 of the License, or *
* (at your option) any later version. *
* *
* You should have received a copy of the GNU General Public License *
* along with this program. If not, see <http://www.gnu.org/licenses/>. *
*************************************************************************/
using System;
using System.Collections.Generic;
using System.Net;
using System.Threading;
using System.Threading.Channels;
using System.Threading.Tasks;
using Server.Logging;
namespace Server.Network.Bans.CrowdSec;
/// <summary>
/// Contributes locally-decided bans to CrowdSec via the LAPI alerts API. Non-blocking on the accept
/// path: <see cref="Report"/> enqueues onto a bounded, drop-on-overflow channel drained by a single
/// background task that coalesces by IP and POSTs batched alerts.
/// </summary>
public sealed class CrowdSecReporter : IBanReporter
{
private static readonly ILogger logger = LogFactory.GetLogger(typeof(CrowdSecReporter));
internal readonly record struct ReportItem(IPAddress Ip, TimeSpan Ttl, string Reason, bool Retract);
// Bounded retry for transient LAPI failures during a drain send (network blips, 5xx). Distinct from
// CrowdSecAlertClient.SendWithRetryAsync's single 401-relogin retry, which is an auth concern.
private static readonly TimeSpan[] _retryDelays = [TimeSpan.FromSeconds(1), TimeSpan.FromSeconds(2)];
private ICrowdSecAlertClient _client;
private CrowdSecSettings _settings;
private Channel<ReportItem> _queue;
private CancellationTokenSource _cts;
private Task _drainTask;
private int _dropped;
private int _sendFailures;
public CrowdSecReporter()
{
}
// Test/embedding ctor with an injected client + settings.
internal CrowdSecReporter(ICrowdSecAlertClient client, CrowdSecSettings settings)
{
_client = client;
_settings = settings;
_queue = CreateQueue(settings.MaxQueue);
}
public string Name => "crowdsec";
public bool CanRetract => true;
public int DroppedCount => _dropped;
/// <summary>
/// Batches ultimately dropped after the bounded transient-retry in <see cref="DrainLoop"/> gave up.
/// Distinct from <see cref="DroppedCount"/> (queue-overflow drops on the accept path): this counts
/// sustained LAPI outages so operators can see contribution loss instead of it being silent.
/// </summary>
public int SendFailureCount => _sendFailures;
/// <summary>The drain loop's task, so tests can assert it stays alive until the loop exits.</summary>
internal Task DrainTaskForTesting => _drainTask;
public static void Configure()
{
BanChannel.Register(new CrowdSecReporter());
}
public void Register()
{
CrowdSecConfiguration.Load();
_settings ??= CrowdSecConfiguration.Settings;
}
public void Start(CancellationToken token)
{
if (!_settings.ReportingEnabled)
{
logger.Information("CrowdSec reporter disabled (machineId/password empty in crowdsec.json)");
return;
}
_client ??= new CrowdSecAlertClient(_settings);
_queue ??= CreateQueue(_settings.MaxQueue);
_cts = CancellationTokenSource.CreateLinkedTokenSource(token);
_drainTask = Task.Run(() => DrainLoop(_cts.Token), _cts.Token);
}
public void Stop()
{
_cts?.Cancel();
// The flush below reads a SingleReader channel, so wait for the drain to actually exit first.
var drainExited = true;
try
{
// Wait(timeout) is false only on timeout; a throw means faulted/cancelled, which is still exited.
drainExited = _drainTask == null || _drainTask.Wait(TimeSpan.FromSeconds(2));
}
catch
{
// Ignored: a faulted wait means the drain has completed and released the channel.
}
_cts?.Dispose();
_cts = null;
if (drainExited)
{
FlushRemainingOnStop();
}
_client?.Dispose();
_drainTask = null;
}
/// <summary>
/// Best-effort bounded flush of whatever is still queued at shutdown. Blocking is correct here — the
/// loop has stopped ticking — but must not happen on the loop thread: <see cref="Stop"/> runs where
/// <c>SynchronizationContext.Current</c> is the <c>EventLoopContext</c>, and a captured continuation
/// would be posted to a queue nothing pumps any more. <see cref="Task.Run(Func{Task})"/> keeps the
/// chain on the pool; the bounded wait caps a wedged send at a few seconds of shutdown.
/// </summary>
private void FlushRemainingOnStop()
{
if (_queue == null || _client == null)
{
return;
}
_queue.Writer.TryComplete();
List<ReportItem> reports = [];
List<ReportItem> retracts = [];
while (_queue.Reader.TryRead(out var item))
{
(item.Retract ? retracts : reports).Add(item);
}
if (reports.Count == 0 && retracts.Count == 0)
{
return;
}
try
{
if (!Task.Run(() => FlushRemainingOnStopAsync(reports, retracts)).Wait(TimeSpan.FromSeconds(4)))
{
logger.Warning(
"CrowdSec flush-on-stop timed out; {Count} item(s) not contributed",
reports.Count + retracts.Count
);
}
}
catch (Exception e)
{
logger.Warning(e, "CrowdSec flush-on-stop failed");
}
}
/// <summary>
/// Uses a fresh token, not the drain loop's already-cancelled one, which would fail every send
/// immediately. Reports go as one deduped batch; retracts go as individual DELETEs so an admin's
/// unban propagates on a clean shutdown. Leftovers self-heal via
/// <see cref="CrowdSecSettings.ManualBanDuration"/>.
/// </summary>
private async Task FlushRemainingOnStopAsync(List<ReportItem> reports, List<ReportItem> retracts)
{
using var flushCts = new CancellationTokenSource(TimeSpan.FromSeconds(3));
if (reports.Count > 0)
{
var alerts = BuildAlerts(reports, _settings, DateTime.UtcNow);
try
{
await _client.PostAlertsAsync(alerts, flushCts.Token).ConfigureAwait(false);
}
catch (Exception e)
{
logger.Warning(e, "CrowdSec flush-on-stop reports failed");
RecordSendFailure(alerts.Count);
}
}
HashSet<string> seen = [];
for (var i = 0; i < retracts.Count; i++)
{
if (flushCts.IsCancellationRequested)
{
break; // out of budget; the rest self-heal via ManualBanDuration
}
var ip = retracts[i].Ip;
if (!seen.Add(ip.ToString()))
{
continue;
}
try
{
await _client.DeleteDecisionsAsync(_settings.Origin, ip, flushCts.Token).ConfigureAwait(false);
}
catch (Exception e)
{
logger.Warning(e, "CrowdSec flush-on-stop retract failed for {Address}", ip);
RecordSendFailure(1);
}
}
}
public void Report(IPAddress address, TimeSpan ttl, string reason) =>
Enqueue(new ReportItem(address, ttl, reason, false));
public void Retract(IPAddress address) =>
Enqueue(new ReportItem(address, TimeSpan.Zero, "retract", true));
private void Enqueue(ReportItem item)
{
if (_queue == null || !_queue.Writer.TryWrite(item))
{
Interlocked.Increment(ref _dropped);
}
}
// FullMode.Wait (the default) makes TryWrite return false immediately when the channel is full
// instead of blocking the caller — exactly the non-blocking drop-on-overflow behavior the accept
// path requires. DropWrite would silently discard the new item and always report success, which
// would make overflow undetectable.
private static Channel<ReportItem> CreateQueue(int capacity) =>
Channel.CreateBounded<ReportItem>(new BoundedChannelOptions(Math.Max(1, capacity))
{
FullMode = BoundedChannelFullMode.Wait,
SingleReader = true
});
// Must return Task: Start() passes this to Task.Run, which has no Func<ValueTask> overload, so a
// ValueTask would bind to Task.Run<TResult> and yield a Task<ValueTask> that completes at the first
// await rather than when the loop exits.
private async Task DrainLoop(CancellationToken token)
{
var reader = _queue.Reader;
while (!token.IsCancellationRequested)
{
try
{
if (!await reader.WaitToReadAsync(token).ConfigureAwait(false))
{
return;
}
// Coalesce a burst before flushing.
await Task.Delay(_settings.FlushInterval, token).ConfigureAwait(false);
List<ReportItem> reports = [];
List<ReportItem> retracts = [];
while (reader.TryRead(out var item))
{
(item.Retract ? retracts : reports).Add(item);
}
if (reports.Count > 0)
{
var alerts = BuildAlerts(reports, _settings, DateTime.UtcNow);
if (!await SendWithBoundedRetryAsync(() => _client.PostAlertsAsync(alerts, token), token)
.ConfigureAwait(false))
{
RecordSendFailure(alerts.Count);
}
}
for (var i = 0; i < retracts.Count; i++)
{
var ip = retracts[i].Ip;
if (!await SendWithBoundedRetryAsync(() => _client.DeleteDecisionsAsync(_settings.Origin, ip, token), token)
.ConfigureAwait(false))
{
RecordSendFailure(1);
}
}
}
catch (OperationCanceledException)
{
return;
}
catch (Exception e)
{
// Contribution is auxiliary: log and keep draining. Never crash the shard.
logger.Warning(e, "CrowdSec contribution flush failed; dropped this batch");
}
}
}
/// <summary>
/// Sends with up to 3 attempts total (1 initial + 2 retries), backing off 1s then 2s between
/// attempts, for transient LAPI failures (network blips, 5xx). Backoff uses <see cref="Task.Delay"/>
/// so it never blocks the thread; a cancellation during backoff propagates as
/// <see cref="OperationCanceledException"/> so the drain loop exits cleanly. Returns false (never
/// throws for a send failure) once attempts are exhausted, so the caller can count the drop and keep
/// draining instead of losing the rest of the batch/queue.
/// </summary>
private static async ValueTask<bool> SendWithBoundedRetryAsync(Func<ValueTask> send, CancellationToken token)
{
for (var attempt = 0; ; attempt++)
{
try
{
await send().ConfigureAwait(false);
return true;
}
catch (OperationCanceledException)
{
throw;
}
catch (Exception e)
{
if (attempt >= _retryDelays.Length)
{
logger.Warning(e, "CrowdSec send failed after {Attempts} attempt(s); giving up", attempt + 1);
return false;
}
var delay = _retryDelays[attempt];
logger.Warning(e, "CrowdSec send failed (attempt {Attempt}); retrying in {Delay}", attempt + 1, delay);
await Task.Delay(delay, token).ConfigureAwait(false);
}
}
}
private void RecordSendFailure(int itemCount)
{
var total = Interlocked.Increment(ref _sendFailures);
logger.Warning(
"CrowdSec contribution batch dropped after retries ({Items} item(s)); total dropped batches: {Total}",
itemCount,
total
);
}
/// <summary>Coalesces items by IP (last write wins) and builds one alert per unique address.</summary>
internal static List<CrowdSecAlert> BuildAlerts(IEnumerable<ReportItem> items, CrowdSecSettings settings, DateTime nowUtc)
{
Dictionary<string, ReportItem> byIp = [];
foreach (var item in items)
{
byIp[item.Ip.ToString()] = item;
}
var timestamp = nowUtc.ToString("yyyy-MM-ddTHH:mm:ss.fffZ");
var alerts = new List<CrowdSecAlert>(byIp.Count);
foreach (var (value, item) in byIp)
{
var ttl = item.Reason == "manual" || item.Ttl <= TimeSpan.Zero ? settings.ManualBanDuration : item.Ttl;
var scenario = $"{settings.Origin}/{item.Reason}";
alerts.Add(new CrowdSecAlert
{
Scenario = scenario,
Message = $"ModernUO {item.Reason} ban for {value}",
StartAt = timestamp,
StopAt = timestamp,
Source = new CrowdSecSource { Scope = "Ip", Value = value },
Decisions =
[
new CrowdSecDecisionDto
{
Origin = settings.Origin,
Type = "ban",
Scope = "Ip",
Value = value,
Duration = FormatDuration(ttl),
Scenario = scenario
}
]
});
}
return alerts;
}
/// <summary>CrowdSec accepts Go durations; whole seconds are unambiguous and sufficient.</summary>
internal static string FormatDuration(TimeSpan ttl)
{
var seconds = (long)ttl.TotalSeconds;
return $"{Math.Max(1, seconds)}s";
}
}