dohertj2/lmxopcua
1
0
Fork 0
Code Issues 2 Pull Requests 1 Actions Packages Projects Releases Wiki Activity

feat: adopt shared ZB.MOM.WW.Health probes (preserve tiers + OtOpcUaCompat policy)

Browse Source
This commit is contained in:
Joseph Doherty
2026-06-01 13:36:28 -04:00
parent 0b99aceacb
commit 1d729fb0f8
4 changed files with 26 additions and 137 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/Server/ZB.MOM.WW.OtOpcUa.Host/Health/AdminRoleLeaderHealthCheck.cs
-39
View File
@@ -1,39 +0,0 @@
using Microsoft.Extensions.Diagnostics.HealthChecks;
using ZB.MOM.WW.OtOpcUa.Commons.Interfaces;
namespace ZB.MOM.WW.OtOpcUa.Host.Health;
/// <summary>
/// Reports Healthy on the admin-role leader, Degraded on a non-leader admin member. Used by
/// the <c>/health/active</c> endpoint so external load balancers can route admin-singleton
/// traffic to the current leader (cookie sessions still work on either node — DataProtection
/// keys are shared).
/// </summary>
public sealed class AdminRoleLeaderHealthCheck : IHealthCheck
{
private readonly IClusterRoleInfo _roleInfo;
/// <summary>Initializes a new instance of the AdminRoleLeaderHealthCheck class.</summary>
/// <param name="roleInfo">The cluster role information provider.</param>
public AdminRoleLeaderHealthCheck(IClusterRoleInfo roleInfo)
{
_roleInfo = roleInfo;
}
/// <summary>Checks the health status of the admin role leader.</summary>
/// <param name="context">The health check context.</param>
/// <param name="cancellationToken">The cancellation token.</param>
/// <returns>A task representing the health check operation.</returns>
public Task<HealthCheckResult> CheckHealthAsync(HealthCheckContext context, CancellationToken cancellationToken = default)
{
if (!_roleInfo.HasRole("admin"))
return Task.FromResult(HealthCheckResult.Healthy("Node does not carry admin role"));
var leader = _roleInfo.RoleLeader("admin");
var isLeader = leader is not null && leader.Value.Equals(_roleInfo.LocalNode);
return Task.FromResult(isLeader
? HealthCheckResult.Healthy($"Admin leader ({_roleInfo.LocalNode})")
: HealthCheckResult.Degraded($"Admin member but not leader (leader={leader?.Value ?? "<unknown>"})"));
}
}
src/Server/ZB.MOM.WW.OtOpcUa.Host/Health/AkkaClusterHealthCheck.cs
-35
View File
@@ -1,35 +0,0 @@
using Akka.Actor;
using Akka.Cluster;
using Microsoft.Extensions.Diagnostics.HealthChecks;
namespace ZB.MOM.WW.OtOpcUa.Host.Health;
public sealed class AkkaClusterHealthCheck : IHealthCheck
{
private readonly ActorSystem _system;
/// <summary>
/// Initializes a new instance of the AkkaClusterHealthCheck class.
/// </summary>
/// <param name="system">The Akka actor system to check cluster health for.</param>
public AkkaClusterHealthCheck(ActorSystem system)
{
_system = system;
}
/// <summary>
/// Checks the health of the Akka cluster asynchronously.
/// </summary>
/// <param name="context">The health check context.</param>
/// <param name="cancellationToken">Cancellation token.</param>
public Task<HealthCheckResult> CheckHealthAsync(HealthCheckContext context, CancellationToken cancellationToken = default)
{
var cluster = Akka.Cluster.Cluster.Get(_system);
var selfUp = cluster.State.Members.Any(m =>
m.Address == cluster.SelfAddress && m.Status == MemberStatus.Up);
return Task.FromResult(selfUp
? HealthCheckResult.Healthy($"Self Up; {cluster.State.Members.Count} member(s)")
: HealthCheckResult.Degraded("Self not yet Up in cluster"));
}
}
src/Server/ZB.MOM.WW.OtOpcUa.Host/Health/DatabaseHealthCheck.cs
-38
View File
@@ -1,38 +0,0 @@
using Microsoft.EntityFrameworkCore;
using Microsoft.Extensions.Diagnostics.HealthChecks;
using ZB.MOM.WW.OtOpcUa.Configuration;
namespace ZB.MOM.WW.OtOpcUa.Host.Health;
public sealed class DatabaseHealthCheck : IHealthCheck
{
private readonly IDbContextFactory<OtOpcUaConfigDbContext> _dbFactory;
/// <summary>
/// Initializes a new instance of the <see cref="DatabaseHealthCheck"/> class.
/// </summary>
/// <param name="dbFactory">The database context factory for the config database.</param>
public DatabaseHealthCheck(IDbContextFactory<OtOpcUaConfigDbContext> dbFactory)
{
_dbFactory = dbFactory;
}
/// <summary>
/// Checks the health of the configuration database.
/// </summary>
/// <param name="context">The health check context.</param>
/// <param name="cancellationToken">Cancellation token.</param>
public async Task<HealthCheckResult> CheckHealthAsync(HealthCheckContext context, CancellationToken cancellationToken = default)
{
try
{
await using var db = await _dbFactory.CreateDbContextAsync(cancellationToken);
await db.Deployments.AsNoTracking().Take(1).ToListAsync(cancellationToken);
return HealthCheckResult.Healthy("ConfigDb reachable");
}
catch (Exception ex)
{
return HealthCheckResult.Unhealthy("ConfigDb unreachable", ex);
}
}
}
src/Server/ZB.MOM.WW.OtOpcUa.Host/Health/HealthEndpoints.cs
+26 -25
View File
@@ -1,25 +1,40 @@
using Microsoft.AspNetCore.Builder;
using Microsoft.AspNetCore.Diagnostics.HealthChecks;
using Microsoft.AspNetCore.Routing;
using Microsoft.EntityFrameworkCore;
using Microsoft.Extensions.DependencyInjection;
using Microsoft.Extensions.Diagnostics.HealthChecks;
using ZB.MOM.WW.Health;
using ZB.MOM.WW.Health.Akka;
using ZB.MOM.WW.Health.EntityFrameworkCore;
using ZB.MOM.WW.OtOpcUa.Configuration;
namespace ZB.MOM.WW.OtOpcUa.Host.Health;
public static class HealthEndpoints
{
/// <summary>
/// Registers the standard ASP.NET Core health-check infrastructure plus the OtOpcUa-specific
/// probes. Mirrors ScadaLink's three-tier pattern: <c>ready</c> = boot ok; <c>active</c> =
/// fully serving traffic; <c>healthz</c> = bare process liveness.
/// Registers the shared ZB.MOM.WW health probes. Tier semantics preserved: configdb + akka on
/// ready+active; admin-leader on active only.
/// </summary>
/// <param name="services">The service collection to register health checks with.</param>
public static IServiceCollection AddOtOpcUaHealth(this IServiceCollection services)
{
services.AddHealthChecks()
.AddCheck<DatabaseHealthCheck>("configdb", tags: new[] { "ready", "active" })
.AddCheck<AkkaClusterHealthCheck>("akka", tags: new[] { "ready", "active" })
.AddCheck<AdminRoleLeaderHealthCheck>("admin-leader", tags: new[] { "active" });
.AddTypeActivatedCheck<DatabaseHealthCheck<OtOpcUaConfigDbContext>>(
"configdb",
failureStatus: null,
tags: new[] { ZbHealthTags.Ready, ZbHealthTags.Active },
args: new DatabaseHealthCheckOptions<OtOpcUaConfigDbContext>
{
ProbeQuery = static (db, ct) => db.Deployments.AsNoTracking().Take(1).ToListAsync(ct),
})
.AddTypeActivatedCheck<AkkaClusterHealthCheck>(
"akka",
failureStatus: null,
tags: new[] { ZbHealthTags.Ready, ZbHealthTags.Active },
args: AkkaClusterStatusPolicy.OtOpcUaCompat)
.AddTypeActivatedCheck<ActiveNodeHealthCheck>(
"admin-leader",
failureStatus: null,
tags: new[] { ZbHealthTags.Active },
args: "admin");
return services;
}
@@ -27,21 +42,7 @@ public static class HealthEndpoints
/// <param name="app">The endpoint route builder.</param>
public static IEndpointRouteBuilder MapOtOpcUaHealth(this IEndpointRouteBuilder app)
{
// AllowAnonymous on all three — Traefik / k8s liveness probes / load-balancers
// hit these without credentials. Without it the AddOtOpcUaAuth fallback policy
// 401s every probe and Traefik marks every backend unhealthy.
app.MapHealthChecks("/health/ready", new HealthCheckOptions
{
Predicate = c => c.Tags.Contains("ready"),
}).AllowAnonymous();
app.MapHealthChecks("/health/active", new HealthCheckOptions
{
Predicate = c => c.Tags.Contains("active"),
}).AllowAnonymous();
app.MapHealthChecks("/healthz", new HealthCheckOptions
{
Predicate = _ => false, // process-liveness only — no probes run.
}).AllowAnonymous();
app.MapZbHealth();
return app;
}
}