Added download yq to fix the issue.

Seperated demo for node pool management
Added overview and link to tutorial
Added comments to CHANGELOG

Signed-off-by: Jason Chua <[email protected]>

.gitignore

@@ -26,3 +26,4 @@
 # Binaries
 vhive
 setup_tool
+go.work.sum

CHANGELOG.md

@@ -4,6 +4,9 @@
 
 ### Added
 
+- Added support for [`OpenYurt`](https://openyurt.io/), an open platform that extends upstream Kubernetes to run on edge node pools. More details on how to the `Knative-atop-OpenYurt` mode are described [here](scripts/openyurt-deployer/README.md).
+
+
 ### Changed
 
 - Removed the utils and examples from the vHive repo, moved to [vSwarm](https://github.com/vhive-serverless/vSwarm).

README.md

@@ -31,7 +31,7 @@ including functions autoscaling and cold-start delay optimization with several s
 
 vHive has added support for the state-of-the-art extension [eStargz](https://github.com/containerd/stargz-snapshotter) to container layers and lazy pull support for container images.
 
-vHive has added support for [`OpenYurt`](https://openyurt.io/), an open platform that extends upstream Kubernetes to the edge. It will assist in node management for cloud and edge nodes and can be deployed with [`openyurt-deployer`](scripts/openyurt-deployer/README.md).
+vHive supports running Knative on top of [`OpenYurt`](https://openyurt.io/), an open platform that extends upstream Kubernetes to run on edge node pools. More details on how to the `Knative-atop-OpenYurt` mode are described [here](scripts/openyurt-deployer/README.md).
 
 ## vHive architecture
 

go.work.sum

@@ -1,4 +1,7 @@
 golang.org/x/crypto v0.16.0 h1:mMMrFzRSCF0GvB7Ne27XVtVAaXLrPmgPC7/v0tkwHaY=
 golang.org/x/crypto v0.16.0/go.mod h1:gCAAfMLgwOJRpTjQ2zCCt2OcSfYMTeZVSRtQlPC7Nq4=
+golang.org/x/crypto v0.18.0 h1:PGVlW0xEltQnzFZ55hkuX5+KLyrMYhHld1YHO4AKcdc=
+golang.org/x/crypto v0.18.0/go.mod h1:R0j02AL6hcrfOiy9T4ZYp/rcWeMxM3L6QYxlOuEG1mg=
 golang.org/x/net v0.10.0/go.mod h1:0qNGK6F8kojg2nk9dLZ2mShWaEBan6FAoqfSigmmuDg=
 golang.org/x/term v0.15.0/go.mod h1:BDl952bC7+uMoWR75FIrCDx79TPU9oHkTZ9yRbYOrX0=
+golang.org/x/term v0.16.0/go.mod h1:yn7UURbUtPyrVJPGPq404EukNFxcm/foM+bV/bfcDsY=

scripts/openyurt-deployer/README.md

@@ -1,25 +1,27 @@
 # Quick set-up `OpenYurt`
 
-## 1.1 Introduction
+![Architecture Diagram](/docs/figures/knative_atop_openyurt.png)
+
+Refer to [Serverless Functions atop of Cloud-Edge Deployment](https://docs.google.com/presentation/d/1QPwwvBhE1dZy430x_wDZPgsUYtlFNlteT9TPk3VbRsY/edit?usp=drive_link) tutorial to understand its application.
 
-This program extends [`EasyOpenyurt`](https://github.com/flyinghorse0510/easy_openyurt) to automate the set up process of an `OpenYurt` cluster. 
+## 1.1 Introduction
 
-It support setting up a Kubernetes cluster using kubeadm and then deploy `OpenYurt` and Knative on it. It is compatible with vHive stock-only mode.
+`OpenYurt` extends Kubernetes from a single cluster to a federation of clusters, offers a seamless way to manage distributed nodes and applications efficiently. This guide outlines how to establish a Kubernetes cluster via kubeadm and then deploy `OpenYurt` and Knative on it, which is compatible with vHive stock-only mode.
 
 ## 1.2 About [`OpenYurt`](https://openyurt.io/docs/#:~:text=Furthermore%2C%20OpenYurt%20enhances%20node%20reliability,node%20heartbeats%20to%20the%20cloud.)
 
 ### Key Features
 
 #### 1. Powerful edge autonomy capability
-`OpenYurt` addresses this issue by implementing a per-node proxy (`YurtHub`) along with local storage to cache the state of the cloud apiserver. Consequently, when a node loses its connection, the cached states remain accessible to Kubelet, `KubeProxy`, and any user Pods.
+`OpenYurt` is designed to ensure that applications running at the edge of the network remain operational, even in poor connectivity conditions. `OpenYurt` addresses this issue by implementing a per-node proxy (`YurtHub`) along with local storage to cache the state of the cloud apiserver. Consequently, when a node loses its connection, the cached states remain accessible to Kubelet, `KubeProxy`, and any user Pods.
 
 #### 2. Cross `NodePool` network communication capability
 
-In an edge computing Kubernetes cluster, nodes are often distributed across various geographical regions. Consequently, when relying on a native Container Network Interface (CNI) solution, Pods within different `NodePools` may be unable to communicate using Pod IP, Service IP, or Node IP, particularly if each `NodePool` resides within its own isolated LAN. Raven offers a networking solution that enables `cross-NodePool` communication within an `OpenYurt` cluster.
+In an edge computing Kubernetes cluster, nodes are often distributed across various geographical regions. The edge nodes fell within same physical region are referred to as `Pools` in `OpenYurt`. Consequently, when relying on a native Container Network Interface (CNI) solution, Pods within different `NodePools` may be unable to communicate using Pod IP, Service IP, or Node IP, particularly if each `NodePool` resides within its own isolated LAN. `OpenYurt` introduces a gateway called `Raven`, which offers a networking solution that enables `cross-NodePool` communication within an `OpenYurt` cluster.
 
 #### 3. `Multi-NodePool` Management
 
-In order to manage applications and traffic in multiple node pools conveniently, `YurtAppSet` and `YurtAppDaemon` are introduced for managing workloads in `multi-nodepool`, and service topology capability for routing traffic in node pool.
+In order to manage applications and traffic in multiple node pools conveniently, `YurtAppSet` and `YurtAppDaemon` are introduced for managing workloads in `multi-nodepool`, and service topology capability (geographical location, network latency, etc.) for routing traffic in node pool.
 
 ## 2. Brief overview
 
@@ -105,6 +107,16 @@ go build .
 ## 4. Demo: Create `NodePool` And Deploy service on it
 **Referenced from [`OpenYurt`](https://openyurt.io/docs/user-manuals/workload/node-pool-management)*
 
+### 4.1 
+
+This will set up the node pool management
+```
+./openyurt_deployer demo
+```
+
+
+### 4.2 
+
 The demo would deploy a helloworld function to cloud node pool or edge node pool
 
 Deploy the demo:

scripts/openyurt-deployer/configurations.go

@@ -98,6 +98,7 @@ type SystemEnvironmentStruct struct {
 	PmuToolsRepoUrl                     string
 	ProtocVersion                       string
 	ProtocDownloadUrlTemplate           string
+	YqDownloadUrl                       string
 }
 
 type YurtEnvironment struct {
@@ -126,8 +127,8 @@ type VHiveConfigStruct struct {
 var Demo = DemoEnvironment{
 	CloudYamlFile:      "cloud.yaml",
 	EdgeYamlFile:       "edge.yaml",
-	CloudBenchYamlFile: "cloud-bench.yaml",
-	EdgeBenchYamlFile:  "edge-bench.yaml",
+	CloudBenchYamlFile: "cloudBench.yaml",
+	EdgeBenchYamlFile:  "edgeBench.yaml",
 	YurtAppSetYamlFile: "yurt.yaml",
 	CloudPoolName:      "cloud",
 	EdgePoolName:       "edge",
@@ -159,6 +160,7 @@ var System = SystemEnvironmentStruct{
 	CurrentDir:          "",
 	UserHomeDir:         "",
 	NodeHostName:        "",
+	YqDownloadUrl:       "https://github.com/mikefarah/yq/releases/latest/download/yq_linux_%s",
 }
 
 var VHive = VHiveConfigStruct{

scripts/openyurt-deployer/main.go

@@ -83,7 +83,9 @@ func main() {
 		deployNodes(*deployerConf)
 	case "knative":
 		deployKnative(*deployerConf)
-	case "demo-deploy-on-edge":
+	case "demo": // this is to add node management properties
+		buildDemo(*deployerConf)
+	case "demo-deploy-on-edge": // to add benchmark
 		demo(*deployerConf, false)
 	case "demo-deploy-on-cloud":
 		demo(*deployerConf, true)
@@ -237,16 +239,25 @@ func deployOpenYurt(deployerConfFile string) {
 	port := lines[1]
 	token := lines[2]
 
+	// 2.1 label your nodes
+	for _, worker := range workerNodes {
+		masterNode.YurtMasterExpand(&worker)
+		utils.InfoPrintf("Master has expanded to worker:%s\n", worker.Configs.System.NodeHostName)
+	}
+	utils.SuccessPrintf("Master has expaned to all nodes!\n")
+
+	// 2.2 Set up up Yurthub & 2.3 Configure Kubelet on Master
 	for _, worker := range workerNodes {
 		worker.YurtWorkerJoin(addr, port, token)
 		utils.InfoPrintf("worker %s joined yurt cluster!\n", worker.Configs.System.NodeHostName)
 	}
 	utils.SuccessPrintf("All nodes joined yurt cluster, start to expand\n")
+
+	// 2.4 Restart pods on edge
 	for _, worker := range workerNodes {
-		masterNode.YurtMasterExpand(&worker)
+		masterNode.YurtRestart(&worker)
 		utils.InfoPrintf("Master has expanded to worker:%s\n", worker.Configs.System.NodeHostName)
 	}
-	utils.SuccessPrintf("Master has expaned to all nodes!\n")
 
 	for _, node := range nodeList {
 		utils.InfoPrintf("node: %s\n", node.Name)
@@ -269,12 +280,26 @@ func demo(deployerConfFile string, isCloud bool) {
 	workerNodes := nodeList[1:]
 
 	// run demo, should only be executed after deployment
-	utils.SuccessPrintf("Start to init demo\n")
+	utils.SuccessPrintf("Start to build cloud and edge nodepools\n")
 	masterNode.Demo(isCloud)
 	utils.SuccessPrintf("Demo finished!\n")
 	masterNode.PrintDemoInfo(workerNodes, isCloud)
 }
 
+func buildDemo(deployerConfFile string) {
+
+	nodesInfo, err := readAndUnMarshall(deployerConfFile)
+	utils.CheckErrorWithMsg(err, "Failed to read and unmarshal deployer configuration JSON")
+	nodeList := initializeNodes(nodesInfo)
+	masterNode := nodeList[0]
+	workerNodes := nodeList[1:]
+
+	// run workload for node pool management
+	utils.SuccessPrintf("Run Build demo\n")
+	masterNode.BuildDemo(workerNodes)
+	utils.SuccessPrintf("Demo finished!\n")
+}
+
 func printDemo(deployerConfFile string) {
 
 	nodesInfo, err := readAndUnMarshall(deployerConfFile)