Phase 3B: Site I/O & Observability — Communication, DCL, Script/Alarm actors, Health, Event Logging
Communication Layer (WP-1–5): - 8 message patterns with correlation IDs, per-pattern timeouts - Central/Site communication actors, transport heartbeat config - Connection failure handling (no central buffering, debug streams killed) Data Connection Layer (WP-6–14, WP-34): - Connection actor with Become/Stash lifecycle (Connecting/Connected/Reconnecting) - OPC UA + LmxProxy adapters behind IDataConnection - Auto-reconnect, bad quality propagation, transparent re-subscribe - Write-back, tag path resolution with retry, health reporting - Protocol extensibility via DataConnectionFactory Site Runtime (WP-15–25, WP-32–33): - ScriptActor/ScriptExecutionActor (triggers, concurrent execution, blocking I/O dispatcher) - AlarmActor/AlarmExecutionActor (ValueMatch/RangeViolation/RateOfChange, in-memory state) - SharedScriptLibrary (inline execution), ScriptRuntimeContext (API) - ScriptCompilationService (Roslyn, forbidden API enforcement, execution timeout) - Recursion limit (default 10), call direction enforcement - SiteStreamManager (per-subscriber bounded buffers, fire-and-forget) - Debug view backend (snapshot + stream), concurrency serialization - Local artifact storage (4 SQLite tables) Health Monitoring (WP-26–28): - SiteHealthCollector (thread-safe counters, connection state) - HealthReportSender (30s interval, monotonic sequence numbers) - CentralHealthAggregator (offline detection 60s, online recovery) Site Event Logging (WP-29–31): - SiteEventLogger (SQLite, 6 event categories, ISO 8601 UTC) - EventLogPurgeService (30-day retention, 1GB cap) - EventLogQueryService (filters, keyword search, keyset pagination) 541 tests pass, zero warnings.
This commit is contained in:
Joseph Doherty
476
src/ScadaLink.DataConnectionLayer/Actors/DataConnectionActor.cs
Normal file
476
src/ScadaLink.DataConnectionLayer/Actors/DataConnectionActor.cs
Normal file
@@ -0,0 +1,476 @@
|
||||
using Akka.Actor;
|
||||
using Akka.Event;
|
||||
using ScadaLink.Commons.Interfaces.Protocol;
|
||||
using ScadaLink.Commons.Messages.DataConnection;
|
||||
using ScadaLink.Commons.Types.Enums;
|
||||
|
||||
namespace ScadaLink.DataConnectionLayer.Actors;
|
||||
|
||||
/// <summary>
|
||||
/// WP-6: Connection actor using Akka.NET Become/Stash pattern for lifecycle state machine.
|
||||
///
|
||||
/// States:
|
||||
/// - Connecting: stash subscribe/write requests; attempts connection
|
||||
/// - Connected: unstash and process all requests
|
||||
/// - Reconnecting: push bad quality for all subscribed tags, stash new requests,
|
||||
/// fixed-interval reconnect
|
||||
///
|
||||
/// WP-9: Auto-reconnect with bad quality on disconnect.
|
||||
/// WP-10: Transparent re-subscribe after reconnection.
|
||||
/// WP-11: Write-back support (synchronous failure to caller, no S&F).
|
||||
/// WP-12: Tag path resolution with retry.
|
||||
/// WP-13: Health reporting (connection status + tag resolution counts).
|
||||
/// WP-14: Subscription lifecycle (register on create, cleanup on stop).
|
||||
/// </summary>
|
||||
public class DataConnectionActor : UntypedActor, IWithStash, IWithTimers
|
||||
{
|
||||
private readonly ILoggingAdapter _log = Context.GetLogger();
|
||||
private readonly string _connectionName;
|
||||
private readonly IDataConnection _adapter;
|
||||
private readonly DataConnectionOptions _options;
|
||||
|
||||
public IStash Stash { get; set; } = null!;
|
||||
public ITimerScheduler Timers { get; set; } = null!;
|
||||
|
||||
/// <summary>
|
||||
/// Active subscriptions: instanceUniqueName → set of tag paths.
|
||||
/// </summary>
|
||||
private readonly Dictionary<string, HashSet<string>> _subscriptionsByInstance = new();
|
||||
|
||||
/// <summary>
|
||||
/// Subscription IDs returned by the adapter: tagPath → subscriptionId.
|
||||
/// </summary>
|
||||
private readonly Dictionary<string, string> _subscriptionIds = new();
|
||||
|
||||
/// <summary>
|
||||
/// Tags whose path resolution failed and are awaiting retry.
|
||||
/// </summary>
|
||||
private readonly HashSet<string> _unresolvedTags = new();
|
||||
|
||||
/// <summary>
|
||||
/// Subscribers: instanceUniqueName → IActorRef (the Instance Actor).
|
||||
/// </summary>
|
||||
private readonly Dictionary<string, IActorRef> _subscribers = new();
|
||||
|
||||
/// <summary>
|
||||
/// Tracks total subscribed and resolved tags for health reporting.
|
||||
/// </summary>
|
||||
private int _totalSubscribed;
|
||||
private int _resolvedTags;
|
||||
|
||||
public DataConnectionActor(
|
||||
string connectionName,
|
||||
IDataConnection adapter,
|
||||
DataConnectionOptions options)
|
||||
{
|
||||
_connectionName = connectionName;
|
||||
_adapter = adapter;
|
||||
_options = options;
|
||||
}
|
||||
|
||||
protected override void PreStart()
|
||||
{
|
||||
_log.Info("DataConnectionActor [{0}] starting in Connecting state", _connectionName);
|
||||
BecomeConnecting();
|
||||
}
|
||||
|
||||
protected override void PostStop()
|
||||
{
|
||||
_log.Info("DataConnectionActor [{0}] stopping — disposing adapter", _connectionName);
|
||||
// Clean up the adapter asynchronously
|
||||
_ = _adapter.DisposeAsync().AsTask();
|
||||
}
|
||||
|
||||
protected override void OnReceive(object message)
|
||||
{
|
||||
// Default handler — should not be reached due to Become
|
||||
Unhandled(message);
|
||||
}
|
||||
|
||||
// ── Connecting State ──
|
||||
|
||||
private void BecomeConnecting()
|
||||
{
|
||||
_log.Info("[{0}] Entering Connecting state", _connectionName);
|
||||
Become(Connecting);
|
||||
Self.Tell(new AttemptConnect());
|
||||
}
|
||||
|
||||
private void Connecting(object message)
|
||||
{
|
||||
switch (message)
|
||||
{
|
||||
case AttemptConnect:
|
||||
HandleAttemptConnect();
|
||||
break;
|
||||
case ConnectResult result:
|
||||
HandleConnectResult(result);
|
||||
break;
|
||||
case SubscribeTagsRequest:
|
||||
case WriteTagRequest:
|
||||
case UnsubscribeTagsRequest:
|
||||
Stash.Stash();
|
||||
break;
|
||||
case GetHealthReport:
|
||||
ReplyWithHealthReport();
|
||||
break;
|
||||
default:
|
||||
Unhandled(message);
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
// ── Connected State ──
|
||||
|
||||
private void BecomeConnected()
|
||||
{
|
||||
_log.Info("[{0}] Entering Connected state", _connectionName);
|
||||
Become(Connected);
|
||||
Stash.UnstashAll();
|
||||
}
|
||||
|
||||
private void Connected(object message)
|
||||
{
|
||||
switch (message)
|
||||
{
|
||||
case SubscribeTagsRequest req:
|
||||
HandleSubscribe(req);
|
||||
break;
|
||||
case UnsubscribeTagsRequest req:
|
||||
HandleUnsubscribe(req);
|
||||
break;
|
||||
case WriteTagRequest req:
|
||||
HandleWrite(req);
|
||||
break;
|
||||
case AdapterDisconnected:
|
||||
HandleDisconnect();
|
||||
break;
|
||||
case RetryTagResolution:
|
||||
HandleRetryTagResolution();
|
||||
break;
|
||||
case GetHealthReport:
|
||||
ReplyWithHealthReport();
|
||||
break;
|
||||
default:
|
||||
Unhandled(message);
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
// ── Reconnecting State ──
|
||||
|
||||
private void BecomeReconnecting()
|
||||
{
|
||||
_log.Warning("[{0}] Entering Reconnecting state", _connectionName);
|
||||
Become(Reconnecting);
|
||||
|
||||
// WP-9: Push bad quality for all subscribed tags on disconnect
|
||||
PushBadQualityForAllTags();
|
||||
|
||||
// Schedule reconnect attempt
|
||||
Timers.StartSingleTimer("reconnect", new AttemptConnect(), _options.ReconnectInterval);
|
||||
}
|
||||
|
||||
private void Reconnecting(object message)
|
||||
{
|
||||
switch (message)
|
||||
{
|
||||
case AttemptConnect:
|
||||
HandleAttemptConnect();
|
||||
break;
|
||||
case ConnectResult result:
|
||||
HandleReconnectResult(result);
|
||||
break;
|
||||
case SubscribeTagsRequest:
|
||||
case WriteTagRequest:
|
||||
Stash.Stash();
|
||||
break;
|
||||
case UnsubscribeTagsRequest req:
|
||||
// Allow unsubscribe even during reconnect (for cleanup on instance stop)
|
||||
HandleUnsubscribe(req);
|
||||
break;
|
||||
case GetHealthReport:
|
||||
ReplyWithHealthReport();
|
||||
break;
|
||||
default:
|
||||
Unhandled(message);
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
// ── Connection Management ──
|
||||
|
||||
private void HandleAttemptConnect()
|
||||
{
|
||||
_log.Debug("[{0}] Attempting connection...", _connectionName);
|
||||
var self = Self;
|
||||
_adapter.ConnectAsync(new Dictionary<string, string>()).ContinueWith(t =>
|
||||
{
|
||||
if (t.IsCompletedSuccessfully)
|
||||
return new ConnectResult(true, null);
|
||||
return new ConnectResult(false, t.Exception?.GetBaseException().Message);
|
||||
}).PipeTo(self);
|
||||
}
|
||||
|
||||
private void HandleConnectResult(ConnectResult result)
|
||||
{
|
||||
if (result.Success)
|
||||
{
|
||||
_log.Info("[{0}] Connection established", _connectionName);
|
||||
BecomeConnected();
|
||||
}
|
||||
else
|
||||
{
|
||||
_log.Warning("[{0}] Connection failed: {1}. Retrying in {2}s",
|
||||
_connectionName, result.Error, _options.ReconnectInterval.TotalSeconds);
|
||||
Timers.StartSingleTimer("reconnect", new AttemptConnect(), _options.ReconnectInterval);
|
||||
}
|
||||
}
|
||||
|
||||
private void HandleReconnectResult(ConnectResult result)
|
||||
{
|
||||
if (result.Success)
|
||||
{
|
||||
_log.Info("[{0}] Reconnected successfully", _connectionName);
|
||||
|
||||
// WP-10: Transparent re-subscribe — re-establish all active subscriptions
|
||||
ReSubscribeAll();
|
||||
|
||||
BecomeConnected();
|
||||
}
|
||||
else
|
||||
{
|
||||
_log.Warning("[{0}] Reconnect failed: {1}. Retrying in {2}s",
|
||||
_connectionName, result.Error, _options.ReconnectInterval.TotalSeconds);
|
||||
Timers.StartSingleTimer("reconnect", new AttemptConnect(), _options.ReconnectInterval);
|
||||
}
|
||||
}
|
||||
|
||||
private void HandleDisconnect()
|
||||
{
|
||||
_log.Warning("[{0}] Adapter reported disconnect", _connectionName);
|
||||
BecomeReconnecting();
|
||||
}
|
||||
|
||||
// ── Subscription Management (WP-14) ──
|
||||
|
||||
private void HandleSubscribe(SubscribeTagsRequest request)
|
||||
{
|
||||
_log.Debug("[{0}] Subscribing {1} tags for instance {2}",
|
||||
_connectionName, request.TagPaths.Count, request.InstanceUniqueName);
|
||||
|
||||
_subscribers[request.InstanceUniqueName] = Sender;
|
||||
|
||||
if (!_subscriptionsByInstance.ContainsKey(request.InstanceUniqueName))
|
||||
_subscriptionsByInstance[request.InstanceUniqueName] = new HashSet<string>();
|
||||
|
||||
var instanceTags = _subscriptionsByInstance[request.InstanceUniqueName];
|
||||
var self = Self;
|
||||
var sender = Sender;
|
||||
|
||||
Task.Run(async () =>
|
||||
{
|
||||
foreach (var tagPath in request.TagPaths)
|
||||
{
|
||||
if (_subscriptionIds.ContainsKey(tagPath))
|
||||
{
|
||||
// Already subscribed — just track for this instance
|
||||
instanceTags.Add(tagPath);
|
||||
continue;
|
||||
}
|
||||
|
||||
try
|
||||
{
|
||||
var subId = await _adapter.SubscribeAsync(tagPath, (path, value) =>
|
||||
{
|
||||
self.Tell(new TagValueReceived(path, value));
|
||||
});
|
||||
_subscriptionIds[tagPath] = subId;
|
||||
instanceTags.Add(tagPath);
|
||||
_totalSubscribed++;
|
||||
_resolvedTags++;
|
||||
}
|
||||
catch (Exception ex)
|
||||
{
|
||||
// WP-12: Tag path resolution failure — mark as unresolved, retry later
|
||||
_unresolvedTags.Add(tagPath);
|
||||
instanceTags.Add(tagPath);
|
||||
_totalSubscribed++;
|
||||
|
||||
self.Tell(new TagResolutionFailed(tagPath, ex.Message));
|
||||
}
|
||||
}
|
||||
|
||||
return new SubscribeTagsResponse(
|
||||
request.CorrelationId, request.InstanceUniqueName, true, null, DateTimeOffset.UtcNow);
|
||||
}).PipeTo(sender);
|
||||
|
||||
// Start tag resolution retry timer if we have unresolved tags
|
||||
if (_unresolvedTags.Count > 0)
|
||||
{
|
||||
Timers.StartPeriodicTimer(
|
||||
"tag-resolution-retry",
|
||||
new RetryTagResolution(),
|
||||
_options.TagResolutionRetryInterval,
|
||||
_options.TagResolutionRetryInterval);
|
||||
}
|
||||
}
|
||||
|
||||
private void HandleUnsubscribe(UnsubscribeTagsRequest request)
|
||||
{
|
||||
_log.Debug("[{0}] Unsubscribing all tags for instance {1}",
|
||||
_connectionName, request.InstanceUniqueName);
|
||||
|
||||
if (!_subscriptionsByInstance.TryGetValue(request.InstanceUniqueName, out var tags))
|
||||
return;
|
||||
|
||||
// WP-14: Cleanup on Instance Actor stop
|
||||
foreach (var tagPath in tags)
|
||||
{
|
||||
// Check if any other instance is still subscribed to this tag
|
||||
var otherSubscribers = _subscriptionsByInstance
|
||||
.Where(kvp => kvp.Key != request.InstanceUniqueName && kvp.Value.Contains(tagPath))
|
||||
.Any();
|
||||
|
||||
if (!otherSubscribers && _subscriptionIds.TryGetValue(tagPath, out var subId))
|
||||
{
|
||||
_ = _adapter.UnsubscribeAsync(subId);
|
||||
_subscriptionIds.Remove(tagPath);
|
||||
_unresolvedTags.Remove(tagPath);
|
||||
_totalSubscribed--;
|
||||
if (!_unresolvedTags.Contains(tagPath))
|
||||
_resolvedTags--;
|
||||
}
|
||||
}
|
||||
|
||||
_subscriptionsByInstance.Remove(request.InstanceUniqueName);
|
||||
_subscribers.Remove(request.InstanceUniqueName);
|
||||
}
|
||||
|
||||
// ── Write Support (WP-11) ──
|
||||
|
||||
private void HandleWrite(WriteTagRequest request)
|
||||
{
|
||||
_log.Debug("[{0}] Writing to tag {1}", _connectionName, request.TagPath);
|
||||
var sender = Sender;
|
||||
|
||||
// WP-11: Write through DCL to device, failure returned synchronously
|
||||
_adapter.WriteAsync(request.TagPath, request.Value).ContinueWith(t =>
|
||||
{
|
||||
if (t.IsCompletedSuccessfully)
|
||||
{
|
||||
var result = t.Result;
|
||||
return new WriteTagResponse(
|
||||
request.CorrelationId, result.Success, result.ErrorMessage, DateTimeOffset.UtcNow);
|
||||
}
|
||||
return new WriteTagResponse(
|
||||
request.CorrelationId, false, t.Exception?.GetBaseException().Message, DateTimeOffset.UtcNow);
|
||||
}).PipeTo(sender);
|
||||
}
|
||||
|
||||
// ── Tag Resolution Retry (WP-12) ──
|
||||
|
||||
private void HandleRetryTagResolution()
|
||||
{
|
||||
if (_unresolvedTags.Count == 0)
|
||||
{
|
||||
Timers.Cancel("tag-resolution-retry");
|
||||
return;
|
||||
}
|
||||
|
||||
_log.Debug("[{0}] Retrying resolution for {1} unresolved tags", _connectionName, _unresolvedTags.Count);
|
||||
|
||||
var self = Self;
|
||||
var toResolve = _unresolvedTags.ToList();
|
||||
|
||||
foreach (var tagPath in toResolve)
|
||||
{
|
||||
_adapter.SubscribeAsync(tagPath, (path, value) =>
|
||||
{
|
||||
self.Tell(new TagValueReceived(path, value));
|
||||
}).ContinueWith(t =>
|
||||
{
|
||||
if (t.IsCompletedSuccessfully)
|
||||
return new TagResolutionSucceeded(tagPath, t.Result) as object;
|
||||
return new TagResolutionFailed(tagPath, t.Exception?.GetBaseException().Message ?? "Unknown error");
|
||||
}).PipeTo(self);
|
||||
}
|
||||
}
|
||||
|
||||
// ── Bad Quality Push (WP-9) ──
|
||||
|
||||
private void PushBadQualityForAllTags()
|
||||
{
|
||||
var now = DateTimeOffset.UtcNow;
|
||||
foreach (var (instanceName, tags) in _subscriptionsByInstance)
|
||||
{
|
||||
if (!_subscribers.TryGetValue(instanceName, out var subscriber))
|
||||
continue;
|
||||
|
||||
subscriber.Tell(new ConnectionQualityChanged(_connectionName, QualityCode.Bad, now));
|
||||
}
|
||||
}
|
||||
|
||||
// ── Re-subscribe (WP-10) ──
|
||||
|
||||
private void ReSubscribeAll()
|
||||
{
|
||||
_log.Info("[{0}] Re-subscribing {1} tags after reconnect", _connectionName, _subscriptionIds.Count);
|
||||
|
||||
var self = Self;
|
||||
var allTags = _subscriptionIds.Keys.ToList();
|
||||
_subscriptionIds.Clear();
|
||||
_resolvedTags = 0;
|
||||
|
||||
foreach (var tagPath in allTags)
|
||||
{
|
||||
_adapter.SubscribeAsync(tagPath, (path, value) =>
|
||||
{
|
||||
self.Tell(new TagValueReceived(path, value));
|
||||
}).ContinueWith(t =>
|
||||
{
|
||||
if (t.IsCompletedSuccessfully)
|
||||
return new TagResolutionSucceeded(tagPath, t.Result) as object;
|
||||
return new TagResolutionFailed(tagPath, t.Exception?.GetBaseException().Message ?? "Unknown error");
|
||||
}).PipeTo(self);
|
||||
}
|
||||
}
|
||||
|
||||
// ── Health Reporting (WP-13) ──
|
||||
|
||||
private void ReplyWithHealthReport()
|
||||
{
|
||||
var status = _adapter.Status;
|
||||
Sender.Tell(new DataConnectionHealthReport(
|
||||
_connectionName, status, _totalSubscribed, _resolvedTags, DateTimeOffset.UtcNow));
|
||||
}
|
||||
|
||||
// ── Internal message handlers for piped async results ──
|
||||
|
||||
private void HandleTagValueReceived(TagValueReceived msg)
|
||||
{
|
||||
// Fan out to all subscribed instances
|
||||
foreach (var (instanceName, tags) in _subscriptionsByInstance)
|
||||
{
|
||||
if (!tags.Contains(msg.TagPath))
|
||||
continue;
|
||||
|
||||
if (_subscribers.TryGetValue(instanceName, out var subscriber))
|
||||
{
|
||||
subscriber.Tell(new TagValueUpdate(
|
||||
_connectionName, msg.TagPath, msg.Value.Value, msg.Value.Quality, msg.Value.Timestamp));
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// ── Internal messages ──
|
||||
|
||||
internal record AttemptConnect;
|
||||
internal record ConnectResult(bool Success, string? Error);
|
||||
internal record AdapterDisconnected;
|
||||
internal record TagValueReceived(string TagPath, TagValue Value);
|
||||
internal record TagResolutionFailed(string TagPath, string Error);
|
||||
internal record TagResolutionSucceeded(string TagPath, string SubscriptionId);
|
||||
internal record RetryTagResolution;
|
||||
public record GetHealthReport;
|
||||
}
|
||||
Reference in New Issue
Block a user