feat(health): show all cluster nodes (online/offline, primary/standby) in health dashboard
Add NodeStatus record, IClusterNodeProvider interface, and AkkaClusterNodeProvider that queries Akka cluster membership for all site-role nodes. HealthReportSender populates ClusterNodes before each report. UI shows a row per node with hostname, Online/Offline badge, and Primary/Standby badge. Falls back to single-node display if ClusterNodes is not populated.
This commit is contained in:
62
src/ScadaLink.Host/Health/AkkaClusterNodeProvider.cs
Normal file
62
src/ScadaLink.Host/Health/AkkaClusterNodeProvider.cs
Normal file
@@ -0,0 +1,62 @@
|
||||
using Akka.Actor;
|
||||
using Akka.Cluster;
|
||||
using ScadaLink.Commons.Messages.Health;
|
||||
using ScadaLink.HealthMonitoring;
|
||||
using ScadaLink.Host.Actors;
|
||||
|
||||
namespace ScadaLink.Host.Health;
|
||||
|
||||
/// <summary>
|
||||
/// Provides cluster node statuses from Akka.NET cluster membership for health reporting.
|
||||
/// </summary>
|
||||
public class AkkaClusterNodeProvider : IClusterNodeProvider
|
||||
{
|
||||
private readonly AkkaHostedService _akkaService;
|
||||
private readonly string _siteRole;
|
||||
|
||||
public AkkaClusterNodeProvider(AkkaHostedService akkaService, string siteRole)
|
||||
{
|
||||
_akkaService = akkaService;
|
||||
_siteRole = siteRole;
|
||||
}
|
||||
|
||||
public IReadOnlyList<NodeStatus> GetClusterNodes()
|
||||
{
|
||||
var system = _akkaService.ActorSystem;
|
||||
if (system == null) return [];
|
||||
|
||||
var cluster = Cluster.Get(system);
|
||||
var selfAddress = cluster.SelfAddress;
|
||||
var leader = cluster.State.Leader;
|
||||
|
||||
var nodes = new List<NodeStatus>();
|
||||
foreach (var member in cluster.State.Members)
|
||||
{
|
||||
if (!member.HasRole(_siteRole))
|
||||
continue;
|
||||
|
||||
var hostname = member.Address.Host ?? member.Address.ToString();
|
||||
var isOnline = member.Status == MemberStatus.Up;
|
||||
var isLeader = member.Address.Equals(leader);
|
||||
var role = isLeader ? "Primary" : "Standby";
|
||||
|
||||
nodes.Add(new NodeStatus(hostname, isOnline, role));
|
||||
}
|
||||
|
||||
// If we have unreachable members, add them as offline
|
||||
foreach (var unreachable in cluster.State.Unreachable)
|
||||
{
|
||||
if (!unreachable.HasRole(_siteRole))
|
||||
continue;
|
||||
|
||||
// Don't duplicate if already in members list
|
||||
if (nodes.Any(n => n.Hostname == (unreachable.Address.Host ?? unreachable.Address.ToString())))
|
||||
continue;
|
||||
|
||||
var hostname = unreachable.Address.Host ?? unreachable.Address.ToString();
|
||||
nodes.Add(new NodeStatus(hostname, false, "Standby"));
|
||||
}
|
||||
|
||||
return nodes;
|
||||
}
|
||||
}
|
||||
@@ -4,6 +4,7 @@ using ScadaLink.DataConnectionLayer;
|
||||
using ScadaLink.ExternalSystemGateway;
|
||||
using ScadaLink.HealthMonitoring;
|
||||
using ScadaLink.Host.Actors;
|
||||
using ScadaLink.Host.Health;
|
||||
using ScadaLink.NotificationService;
|
||||
using ScadaLink.SiteEventLogging;
|
||||
using ScadaLink.SiteRuntime;
|
||||
@@ -42,6 +43,15 @@ public static class SiteServiceRegistration
|
||||
services.AddSingleton<AkkaHostedService>();
|
||||
services.AddHostedService(sp => sp.GetRequiredService<AkkaHostedService>());
|
||||
|
||||
// Cluster node status provider for health reports
|
||||
services.AddSingleton<IClusterNodeProvider>(sp =>
|
||||
{
|
||||
var akkaService = sp.GetRequiredService<AkkaHostedService>();
|
||||
var nodeOptions = sp.GetRequiredService<Microsoft.Extensions.Options.IOptions<NodeOptions>>().Value;
|
||||
var siteRole = $"site-{nodeOptions.SiteId}";
|
||||
return new AkkaClusterNodeProvider(akkaService, siteRole);
|
||||
});
|
||||
|
||||
// Options binding
|
||||
BindSharedOptions(services, config);
|
||||
services.Configure<SiteRuntimeOptions>(config.GetSection("ScadaLink:SiteRuntime"));
|
||||
|
||||
Reference in New Issue
Block a user