From 8224fde3d6c2f69ca1d9d469cf84dabc01a65216 Mon Sep 17 00:00:00 2001
From: xdkaine <55013938+xdkaine@users.noreply.github.com>
Date: Tue, 17 Feb 2026 16:02:15 -0800
Subject: [PATCH 1/2] trying to fix race condition in a more optimal way

---
 internal/proxmox/types.go | 1 +
 internal/proxmox/vms.go   | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/internal/proxmox/types.go b/internal/proxmox/types.go
index 245e505..6a98323 100644
--- a/internal/proxmox/types.go
+++ b/internal/proxmox/types.go
@@ -108,6 +108,7 @@ type ProxmoxNodeStatus struct {
 }
 
 type VirtualResourceConfig struct {
+	Name     string `json:"name"`
 	HardDisk string `json:"scsi0"`
 	Lock     string `json:"lock"`
 	Net0     string `json:"net0"`
diff --git a/internal/proxmox/vms.go b/internal/proxmox/vms.go
index 18fb991..53d95bc 100644
--- a/internal/proxmox/vms.go
+++ b/internal/proxmox/vms.go
@@ -181,7 +181,7 @@ func (s *ProxmoxService) WaitForDisk(node string, vmID int, maxWait time.Duratio
 
 		log.Printf("%+v", configResp)
 
-		if configResp.HardDisk != "" {
+		if configResp.HardDisk != "" && configResp.Name != "" {
 			log.Printf("/nodes/%s/storage/%s/content?vmid=%d", s.Config.Nodes[0], s.Config.StorageID, vmID)
 
 			pendingReq := tools.ProxmoxAPIRequest{

From 5b4e2f10d669613620b120e405a8caaeee80c136 Mon Sep 17 00:00:00 2001
From: xdkaine <55013938+xdkaine@users.noreply.github.com>
Date: Tue, 17 Feb 2026 16:20:38 -0800
Subject: [PATCH 2/2] attempting to fix race conditions part 2

---
 internal/cloning/cloning_service.go | 28 ++++++++++++++++++++--------
 1 file changed, 20 insertions(+), 8 deletions(-)

diff --git a/internal/cloning/cloning_service.go b/internal/cloning/cloning_service.go
index 23b1342..c97a8b4 100644
--- a/internal/cloning/cloning_service.go
+++ b/internal/cloning/cloning_service.go
@@ -270,7 +270,23 @@ func (cs *CloningService) CloneTemplate(req CloneRequest) error {
 	// Release the vmid allocation mutex now that all of the VMs are cloned on proxmox
 	cs.vmidMutex.Unlock()
 
-	// 9. Configure VNet of all VMs
+	// 9. Wait for all router disks to be fully available before configuring VNets.
+	// Proxmox clone is two-phase: the clone lock (Phase 1) releases before the storage
+	// backend finishes writing the disk (Phase 2). If SetPodVnet runs before Phase 2
+	// completes, Proxmox's disk finalization can overwrite the net1 config change,
+	// leaving the router connected to the wrong vnet.
+	log.Printf("Waiting for router disks to be available before configuring VNets")
+	routerDiskReady := make(map[int]bool)
+	for _, routerInfo := range clonedRouters {
+		log.Printf("Waiting for router disk to be available for %s (VMID: %d)", routerInfo.TargetName, routerInfo.VMID)
+		if err := cs.ProxmoxService.WaitForDisk(routerInfo.Node, routerInfo.VMID, cs.Config.RouterWaitTimeout); err != nil {
+			errors = append(errors, fmt.Sprintf("router disk unavailable for %s: %v", routerInfo.TargetName, err))
+		} else {
+			routerDiskReady[routerInfo.VMID] = true
+		}
+	}
+
+	// 10. Configure VNet of all VMs
 	log.Printf("Configuring VNets for %d targets", len(req.Targets))
 	for _, target := range req.Targets {
 		vnetName := fmt.Sprintf("kamino%d", target.PodNumber)
@@ -281,7 +297,7 @@ func (cs *CloningService) CloneTemplate(req CloneRequest) error {
 		}
 	}
 
-	// 10. Start all routers and wait for them to be running
+	// 11. Start all routers and wait for them to be running
 	req.SSE.Send(
 		ProgressMessage{
 			Message:  "Starting routers",
@@ -290,11 +306,7 @@ func (cs *CloningService) CloneTemplate(req CloneRequest) error {
 	)
 	log.Printf("Starting %d routers", len(clonedRouters))
 	for _, routerInfo := range clonedRouters {
-		// Wait for router disk to be available
-		log.Printf("Waiting for router disk to be available for %s (VMID: %d)", routerInfo.TargetName, routerInfo.VMID)
-		err = cs.ProxmoxService.WaitForDisk(routerInfo.Node, routerInfo.VMID, cs.Config.RouterWaitTimeout)
-		if err != nil {
-			errors = append(errors, fmt.Sprintf("router disk unavailable for %s: %v", routerInfo.TargetName, err))
+		if !routerDiskReady[routerInfo.VMID] {
 			continue
 		}
 
@@ -314,7 +326,7 @@ func (cs *CloningService) CloneTemplate(req CloneRequest) error {
 		}
 	}
 
-	// 11. Configure all pod routers (separate step after all routers are running)
+	// 12. Configure all pod routers (separate step after all routers are running)
 	req.SSE.Send(
 		ProgressMessage{
 			Message:  "Configuring pod routers",