From 6881e5c53cb2b596dd015d835e2867a47fcc65c1 Mon Sep 17 00:00:00 2001
From: "goodolclint-claude[bot]"
 <323206664+goodolclint-claude[bot]@users.noreply.github.com>
Date: Tue, 1 Sep 2026 06:20:55 +0000
Subject: [PATCH] feat: add ClusterConfigService.WaitForQuorum

Polls GET /cluster/status for the cluster entry with quorate = 1, bounded
by a timeout (60 s default) and tolerating PveApiException while pmxcfs
and corosync restart during cluster formation.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
---
 .../Services/ClusterConfigService.cs          | 47 +++++++++++++++++++
 1 file changed, 47 insertions(+)

diff --git a/src/PSProxmoxVE.Core/Services/ClusterConfigService.cs b/src/PSProxmoxVE.Core/Services/ClusterConfigService.cs
index a66f603..e288816 100644
--- a/src/PSProxmoxVE.Core/Services/ClusterConfigService.cs
+++ b/src/PSProxmoxVE.Core/Services/ClusterConfigService.cs
@@ -1,8 +1,10 @@
 using System;
 using System.Collections.Generic;
+using System.Threading;
 using Newtonsoft.Json.Linq;
 using PSProxmoxVE.Core.Authentication;
 using PSProxmoxVE.Core.Client;
+using PSProxmoxVE.Core.Exceptions;
 using PSProxmoxVE.Core.Models.Cluster;
 using PSProxmoxVE.Core.Utilities;
 
@@ -16,6 +18,9 @@ namespace PSProxmoxVE.Core.Services
     {
         private readonly IPveHttpClient? _injectedClient;
 
+        private static readonly TimeSpan DefaultQuorumTimeout = TimeSpan.FromSeconds(60);
+        private static readonly TimeSpan QuorumPollInterval = TimeSpan.FromSeconds(2);
+
         /// <summary>
         /// Initializes a new instance of the <see cref="ClusterConfigService"/> class.
         /// </summary>
@@ -378,6 +383,48 @@ namespace PSProxmoxVE.Core.Services
             }
         }
 
+        /// <summary>
+        /// Blocks until the cluster reports quorum (GET /cluster/status, quorate = 1).
+        /// </summary>
+        /// <param name="session">The authenticated PVE session.</param>
+        /// <param name="timeout">Maximum time to wait. Defaults to 60 seconds.</param>
+        /// <remarks>
+        /// The cluster-create task completes before corosync converges; until the node
+        /// is quorate it rejects joins with "cluster not ready - no quorum?". API errors
+        /// during that window are transient and are retried until the deadline.
+        /// </remarks>
+        /// <exception cref="TimeoutException">Quorum was not reached before the deadline.</exception>
+        public void WaitForQuorum(PveSession session, TimeSpan? timeout = null)
+        {
+            if (session == null) throw new ArgumentNullException(nameof(session));
+
+            var effectiveTimeout = timeout ?? DefaultQuorumTimeout;
+            var deadline = DateTime.UtcNow.Add(effectiveTimeout);
+
+            while (true)
+            {
+                try
+                {
+                    foreach (var entry in GetClusterStatus(session))
+                    {
+                        if (string.Equals(entry.Type, "cluster", StringComparison.OrdinalIgnoreCase)
+                            && entry.Quorate == 1)
+                            return;
+                    }
+                }
+                catch (PveApiException)
+                {
+                    // pmxcfs and corosync restart while the cluster forms.
+                }
+
+                if (DateTime.UtcNow >= deadline)
+                    throw new TimeoutException(
+                        $"Cluster did not reach quorum within {effectiveTimeout.TotalSeconds:0} seconds.");
+
+                Thread.Sleep(QuorumPollInterval);
+            }
+        }
+
         /// <summary>
         /// Returns the next available VM/CT ID (GET /cluster/nextid).
         /// </summary>