hub instance state refactor (#892)

- Improvement: Provisioning a hibernated instance now automatically wakes it instead of requiring an explicit wake call first
- Improvement: Deprovisioning now accepts instances in Crashed or Hibernated states, not just Provisioned
- Improvement: Added `--consul-health-interval-seconds` and `--consul-deregister-after-seconds` options to control Consul health check behavior (defaults: 10s and 30s)
- Improvement: Consul registration now occurs when provisioning starts; health check intervals are applied once provisioning completes

1 files changed, 28 insertions, 5 deletions

diff --git a/src/zenserver/hub/zenhubserver.cpp b/src/zenserver/hub/zenhubserver.cpp
index 4d70fbddd..269de28c2 100644
--- a/src/zenserver/hub/zenhubserver.cpp
+++ b/src/zenserver/hub/zenhubserver.cpp
@@ -74,6 +74,20 @@ ZenHubServerConfigurator::AddCliOptions(cxxopts::Options& Options)
 
 	Options.add_option("hub",
 					   "",
+					   "consul-health-interval-seconds",
+					   "Interval in seconds between Consul health checks",
+					   cxxopts::value<uint32_t>(m_ServerOptions.ConsulHealthIntervalSeconds)->default_value("10"),
+					   "<seconds>");
+
+	Options.add_option("hub",
+					   "",
+					   "consul-deregister-after-seconds",
+					   "Seconds after which Consul deregisters an unhealthy service",
+					   cxxopts::value<uint32_t>(m_ServerOptions.ConsulDeregisterAfterSeconds)->default_value("30"),
+					   "<seconds>");
+
+	Options.add_option("hub",
+					   "",
 					   "hub-base-port-number",
 					   "Base port number for provisioned instances",
 					   cxxopts::value<uint16_t>(m_ServerOptions.HubBasePortNumber)->default_value("21000"),
@@ -181,7 +195,8 @@ ZenHubServer::OnModuleStateChanged(std::string_view					 HubInstanceId,
 	{
 		return;
 	}
-	if (NewState == HubInstanceState::Provisioned)
+
+	if (NewState == HubInstanceState::Provisioning || NewState == HubInstanceState::Provisioned)
 	{
 		consul::ServiceRegistrationInfo ServiceInfo{
 			.ServiceId				= std::string(ModuleId),
@@ -191,8 +206,12 @@ ZenHubServer::OnModuleStateChanged(std::string_view					 HubInstanceId,
 			.Tags					= std::vector<std::pair<std::string, std::string>>{std::make_pair("module", std::string(ModuleId)),
 																	   std::make_pair("zen-hub", std::string(HubInstanceId)),
 																	   std::make_pair("version", std::string(ZEN_CFG_VERSION))},
-			.HealthIntervalSeconds	= 10,
-			.DeregisterAfterSeconds = 30};
+			.HealthIntervalSeconds	= NewState == HubInstanceState::Provisioning
+										  ? 0u
+										  : m_ConsulHealthIntervalSeconds,	// Disable health checks while not finished provisioning
+			.DeregisterAfterSeconds = NewState == HubInstanceState::Provisioning
+										  ? 0u
+										  : m_ConsulDeregisterAfterSeconds};  // Disable health checks while not finished provisioning
 
 		if (!m_ConsulClient->RegisterService(ServiceInfo))
 		{
@@ -219,7 +238,7 @@ ZenHubServer::OnModuleStateChanged(std::string_view					 HubInstanceId,
 			ZEN_INFO("Deregistered storage server instance for module '{}' at port {} from Consul", ModuleId, Info.Port);
 		}
 	}
-	// Transitional states (Provisioning, Deprovisioning, Hibernating, Waking, Recovering, Crashed)
+	// Transitional states (Deprovisioning, Hibernating, Waking, Recovering, Crashed)
 	// and Hibernated are intentionally ignored.
 }
 
@@ -385,7 +404,9 @@ ZenHubServer::InitializeConsulRegistration(const ZenHubServerConfig& ServerConfi
 
 	try
 	{
-		m_ConsulClient = std::make_unique<consul::ConsulClient>(ServerConfig.ConsulEndpoint, ConsulAccessToken);
+		m_ConsulClient				   = std::make_unique<consul::ConsulClient>(ServerConfig.ConsulEndpoint, ConsulAccessToken);
+		m_ConsulHealthIntervalSeconds  = ServerConfig.ConsulHealthIntervalSeconds;
+		m_ConsulDeregisterAfterSeconds = ServerConfig.ConsulDeregisterAfterSeconds;
 
 		consul::ServiceRegistrationInfo Info;
 		Info.ServiceId	 = fmt::format("zen-hub-{}", ServerConfig.InstanceId);
@@ -399,6 +420,8 @@ ZenHubServer::InitializeConsulRegistration(const ZenHubServerConfig& ServerConfi
 			  std::make_pair("base-port-number", fmt::format("{}", ServerConfig.HubBasePortNumber)),
 			  std::make_pair("instance-limit", fmt::format("{}", ServerConfig.HubInstanceLimit)),
 			  std::make_pair("use-job-object", fmt::format("{}", ServerConfig.HubUseJobObject))};
+		Info.HealthIntervalSeconds	= ServerConfig.ConsulHealthIntervalSeconds;
+		Info.DeregisterAfterSeconds = ServerConfig.ConsulDeregisterAfterSeconds;
 
 		m_ConsulRegistration = std::make_unique<consul::ServiceRegistration>(m_ConsulClient.get(), Info);