Update to add resize cluster support & Bug fixes

1. Add resize cluster support.
2. Fix failed to edit cluster name.
3. Fix failed to edit security group.
4. Update to handle cluster upgrade failed properly.

Dockerfile.dapper

@@ -1,11 +1,8 @@
-FROM registry.suse.com/bci/golang:1.20
+FROM registry.suse.com/bci/golang:1.21
 
 ARG DAPPER_HOST_ARCH
 ENV ARCH=${DAPPER_HOST_ARCH}
 
-ARG HTTP_PROXY
-ARG HTTPS_PROXY
-ENV HTTP_PROXY=${HTTP_PROXY} HTTPS_PROXY=${HTTPS_PROXY}
 RUN zypper ref && \
     zypper -n up && \
     zypper -n in vim wget git tar gzip && \

charts/cce-operator-crd/templates/crds.yaml

@@ -360,6 +360,9 @@ spec:
               phase:
                 nullable: true
                 type: string
+              resizeClusterJobID:
+                nullable: true
+                type: string
               upgradeClusterTaskID:
                 nullable: true
                 type: string

examples/docs/parameters-zh_CN.md

@@ -162,10 +162,21 @@
     "name": "example-update", // 集群名称
     "huaweiCredentialSecret": "cattle-global-data:cc-xxxxx", // 更新云凭证
     "description": "", // 更新集群描述
-    // "version": "v1.25" // 集群升级（暂不支持）
+    // 集群升级，直接改动集群版本即可升级。例如将 1.25 改为 1.27，即升级至 1.27 版本。
+    // 需要注意集群升级前需要手动检查所有插件必须运行正常（没有未就绪的工作负载），否则会造成升级失败。
+    "version": "v1.25",
     "hostNetwork": {
-        "securityGroup": "SECURITY-GROUP-ID" // Security Group
+        "securityGroup": "SECURITY-GROUP-ID" // 修改节点默认安全组
     },
+    // 变更集群 (Resize) 存在约束限制：https://support.huaweicloud.com/usermanual-cce/cce_10_0403.html
+    // 需要额外注意以下几点：
+    // 1. 变更集群规格不支持修改控制节点数量。例如无法将 s1 修改为 s2
+    // 2. 不支持降低集群规格。例如不能将 medium 降为 small
+    // 3. 单控制节点的集群不允许变更到 1000 节点及以上。例如 s1.small (50 节点) 只能升级到 s1.medium (200 节点)。
+    // Resize 请求可能遇到规格不足：Insufficient resources under expected master specifications 错误。
+    "flavor": "cce.s1.medium", // s1：单控制节点CCE集群。
+                              // s2：多控制节点CCE集群 （高可用）。
+                              // small (最大 50 节点), medium (200 节点), large (1k 节点), xlarge (2k 节点)
     "nodePools": [ // 可编辑节点池数组，用于增加/删除节点池
         {
             "name": "nodepool-1", // 节点池名称，可编辑

go.mod

@@ -1,19 +1,19 @@
 module github.com/cnrancher/cce-operator
 
-go 1.19
+go 1.21
 
 require (
 	github.com/Masterminds/semver/v3 v3.2.1
 	github.com/antonfisher/nested-logrus-formatter v1.3.1
-	github.com/huaweicloud/huaweicloud-sdk-go-v3 v0.1.57
+	github.com/huaweicloud/huaweicloud-sdk-go-v3 v0.1.69
 	github.com/rancher/lasso v0.0.0-20230830164424-d684fdeb6f29
 	github.com/rancher/wrangler v1.1.1
 	github.com/rancher/wrangler-api v0.6.1-0.20200427172631-a7c2f09b783e
 	github.com/sirupsen/logrus v1.9.3
 	github.com/stretchr/testify v1.8.4
-	k8s.io/api v0.27.6
-	k8s.io/apimachinery v0.27.6
-	k8s.io/client-go v0.27.6
+	k8s.io/api v0.27.8
+	k8s.io/apimachinery v0.27.8
+	k8s.io/client-go v0.27.8
 )
 
 require (
@@ -48,17 +48,18 @@ require (
 	github.com/prometheus/common v0.32.1 // indirect
 	github.com/prometheus/procfs v0.7.3 // indirect
 	github.com/spf13/pflag v1.0.5 // indirect
+	github.com/tjfoc/gmsm v1.4.1 // indirect
 	go.mongodb.org/mongo-driver v1.12.0 // indirect
-	golang.org/x/crypto v0.11.0 // indirect
-	golang.org/x/mod v0.9.0 // indirect
-	golang.org/x/net v0.12.0 // indirect
+	golang.org/x/crypto v0.14.0 // indirect
+	golang.org/x/mod v0.12.0 // indirect
+	golang.org/x/net v0.17.0 // indirect
 	golang.org/x/oauth2 v0.10.0 // indirect
-	golang.org/x/sync v0.1.0 // indirect
-	golang.org/x/sys v0.10.0 // indirect
-	golang.org/x/term v0.10.0 // indirect
-	golang.org/x/text v0.11.0 // indirect
+	golang.org/x/sync v0.3.0 // indirect
+	golang.org/x/sys v0.13.0 // indirect
+	golang.org/x/term v0.13.0 // indirect
+	golang.org/x/text v0.13.0 // indirect
 	golang.org/x/time v0.3.0 // indirect
-	golang.org/x/tools v0.7.0 // indirect
+	golang.org/x/tools v0.12.0 // indirect
 	google.golang.org/appengine v1.6.7 // indirect
 	google.golang.org/protobuf v1.31.0 // indirect
 	gopkg.in/inf.v0 v0.9.1 // indirect

pkg/apis/cce.pandaria.io/v1/types.go

@@ -81,6 +81,7 @@ type CCEClusterConfigStatus struct {
 	CreatedSNatRuleEIPID string `json:"createdSNatRuleEIPID"` // EIP ID for SNAT Rule
 	CreatedSNATRuleID    string `json:"createdSNATRuleID"`    // SNAT Rule ID
 
+	ResizeClusterJobID   string `json:"resizeClusterJobID"`   // resize cluster job ID
 	UpgradeClusterTaskID string `json:"upgradeClusterTaskID"` // upgrade cluster task ID
 }
 

pkg/controller/cce-cluster-config-handler.go

@@ -6,6 +6,7 @@ import (
 	"net/url"
 	"time"
 
+	"github.com/Masterminds/semver/v3"
 	ccev1 "github.com/cnrancher/cce-operator/pkg/apis/cce.pandaria.io/v1"
 	ccecontrollers "github.com/cnrancher/cce-operator/pkg/generated/controllers/cce.pandaria.io/v1"
 	"github.com/cnrancher/cce-operator/pkg/huawei"
@@ -630,8 +631,29 @@ func (h *Handler) checkAndUpdate(config *ccev1.CCEClusterConfig) (*ccev1.CCEClus
 		return config, err
 	}
 
+	// Get cluster status.
+	cluster, err := cce.ShowCluster(driver.CCE, config.Spec.ClusterID)
+	if err != nil {
+		return config, err
+	}
+	if cluster == nil || cluster.Status == nil || cluster.Spec == nil || cluster.Spec.HostNetwork == nil {
+		return config, fmt.Errorf("GetCluster returns invalid data")
+	}
+
 	// Check cluster upgrade status.
 	if config.Status.UpgradeClusterTaskID != "" {
+		ok, err := clusterUpgradeable(config.Spec.Version, *cluster.Spec.Version)
+		if err != nil {
+			return config, err
+		}
+		if !ok {
+			// Delete UpgradeClusterTaskID if the config cluster version match
+			// the upstream cluster version.
+			config = config.DeepCopy()
+			config.Status.UpgradeClusterTaskID = ""
+			return h.cceCC.UpdateStatus(config)
+		}
+
 		res, err := cce.ShowUpgradeClusterTask(driver.CCE, config.Spec.ClusterID, config.Status.UpgradeClusterTaskID)
 		if err != nil {
 			hwerr, _ := huawei.NewHuaweiError(err)
@@ -649,11 +671,14 @@ func (h *Handler) checkAndUpdate(config *ccev1.CCEClusterConfig) (*ccev1.CCEClus
 				logrus.WithFields(logrus.Fields{
 					"cluster": config.Name,
 					"phase":   config.Status.Phase,
-				}).Infof("cluster [%s] finished upgrade",
-					config.Spec.Name)
+				}).Infof("cluster [%s] upgrade to [%v]",
+					config.Spec.Name, utils.Value(cluster.Spec.Version))
 				config = config.DeepCopy()
 				config.Status.UpgradeClusterTaskID = ""
 				return h.cceCC.UpdateStatus(config)
+			case "Failed":
+				return config, fmt.Errorf("failed to upgrade cluster [%s] to %v, status [%s]",
+					config.Spec.Name, config.Spec.Version, utils.Value(res.Status.Phase))
 			default:
 				logrus.WithFields(logrus.Fields{
 					"cluster": config.Name,
@@ -667,13 +692,6 @@ func (h *Handler) checkAndUpdate(config *ccev1.CCEClusterConfig) (*ccev1.CCEClus
 	}
 
 	// Check cluster status.
-	cluster, err := cce.ShowCluster(driver.CCE, config.Spec.ClusterID)
-	if err != nil {
-		return config, err
-	}
-	if cluster == nil || cluster.Status == nil || cluster.Spec == nil || cluster.Spec.HostNetwork == nil {
-		return config, fmt.Errorf("GetCluster returns invalid data")
-	}
 	switch utils.Value(cluster.Status.Phase) {
 	case cce.ClusterStatusDeleting,
 		cce.ClusterStatusResizing,
@@ -816,8 +834,10 @@ func (h *Handler) updateUpstreamClusterState(
 		return config, nil
 	}
 
-	// Update security group ID for created cluster.
-	if config.Spec.HostNetwork.SecurityGroup != upstreamSpec.HostNetwork.SecurityGroup {
+	// Init security group ID for created cluster if the security group wasn't
+	// provided when creating the cluster.
+	if config.Spec.HostNetwork.SecurityGroup == "" &&
+		config.Spec.HostNetwork.SecurityGroup != upstreamSpec.HostNetwork.SecurityGroup {
 		configUpdate := config.DeepCopy()
 		configUpdate.Spec.HostNetwork.SecurityGroup = upstreamSpec.HostNetwork.SecurityGroup
 		config, err = h.cceCC.Update(configUpdate)
@@ -832,6 +852,57 @@ func (h *Handler) updateUpstreamClusterState(
 	} else if ok {
 		return h.upgradeCluster(config)
 	}
+	// Check cluster flavor is resizable.
+	var clusterResizable = false
+	if config.Spec.Flavor != "" && config.Spec.Flavor != upstreamSpec.Flavor {
+		cv, err := semver.NewVersion(config.Spec.Version)
+		if err != nil {
+			return config, err
+		}
+		minVersion := semver.New(1, 15, 0, "", "")
+		if cv.Compare(minVersion) >= 0 {
+			clusterResizable = true
+		}
+	}
+	if clusterResizable {
+		logrus.WithFields(logrus.Fields{
+			"cluster": config.Name,
+			"phase":   config.Status.Phase,
+		}).Infof("cluster [%s] flavor change detected: %v -> %v",
+			config.Spec.Name, upstreamSpec.Flavor, config.Spec.Flavor)
+
+		res, err := cce.ResizeCluster(
+			driver.CCE,
+			config.Spec.ClusterID,
+			config.Spec.Flavor,
+			config.Spec.ExtendParam.IsAutoPay,
+		)
+		if err != nil {
+			return config, err
+		}
+		if res == nil || res.JobID == nil {
+			return config, fmt.Errorf("ResizeCluster returns invalid data")
+		}
+		logrus.WithFields(logrus.Fields{
+			"cluster": config.Name,
+			"phase":   config.Status.Phase,
+		}).Infof("start resize cluster [%s] job ID %q",
+			config.Spec.Name, utils.Value(res.JobID))
+		err = retry.RetryOnConflict(retry.DefaultRetry, func() error {
+			config, err = h.cceCC.Get(config.Namespace, config.Name, metav1.GetOptions{})
+			if err != nil {
+				return err
+			}
+			configUpdate := config.DeepCopy()
+			configUpdate.Status.ResizeClusterJobID = utils.Value(res.JobID)
+			config, err = h.cceCC.UpdateStatus(configUpdate)
+			return err
+		})
+		if err != nil {
+			return config, err
+		}
+		return h.enqueueUpdate(config)
+	}
 
 	// Update cluster info.
 	if _, err = cce.UpdateCluster(driver.CCE, config); err != nil {

pkg/controller/upstream.go

@@ -41,6 +41,10 @@ func BuildUpstreamClusterState(
 		KubeProxyMode:          c.Spec.KubeProxyMode.Value(),
 		PublicAccess:           false,
 	}
+	if utils.Value(c.Metadata.Alias) != "" && spec.Name != utils.Value(c.Metadata.Alias) {
+		// Set cluster name to edited alias instead of the original name.
+		spec.Name = utils.Value(c.Metadata.Alias)
+	}
 	if c.Spec.HostNetwork != nil {
 		spec.HostNetwork.VpcID = c.Spec.HostNetwork.Vpc
 		spec.HostNetwork.SubnetID = c.Spec.HostNetwork.Subnet

pkg/huawei/cce/cce.go

@@ -261,6 +261,25 @@ func ShowUpgradeClusterTask(
 	return res, err
 }
 
+func ResizeCluster(
+	client *cce.CceClient, ID, flavor, isAutoPay string,
+) (*model.ResizeClusterResponse, error) {
+	req := &model.ResizeClusterRequest{
+		ClusterId: ID,
+		Body: &model.ResizeClusterRequestBody{
+			FlavorResize: flavor,
+			ExtendParam: &model.ResizeClusterRequestExtendParam{
+				IsAutoPay: &isAutoPay,
+			},
+		},
+	}
+	res, err := client.ResizeCluster(req)
+	if err != nil {
+		logrus.Debugf("ResizeCluster failed: %v", utils.PrintObject(req))
+	}
+	return res, err
+}
+
 func DeleteCluster(client *cce.CceClient, ID string) (*model.DeleteClusterResponse, error) {
 	res, err := client.DeleteCluster(&model.DeleteClusterRequest{
 		ClusterId: ID,