Merge pull request #4512 from cyclinder/0.8_cp/coordinator_detect_tim…

…eout1

coodirnator: optimize the detectiong timeout for ip conflict and gateway detection

Makefile

@@ -278,7 +278,7 @@ unittest-tests: check_test_label
 	@echo "run unittest-tests"
 	$(QUIET) $(ROOT_DIR)/tools/scripts/ginkgo.sh \
 		--cover --coverprofile=./coverage.out --covermode set \
-		--json-report unittestreport.json --label-filter $(E2E_GINKGO_UTLABELS) \
+		--json-report unittestreport.json  \
 		-randomize-suites -randomize-all --keep-going  --timeout=1h  -p \
 		-vv  -r $(ROOT_DIR)/pkg $(ROOT_DIR)/cmd
 	$(QUIET) go tool cover -html=./coverage.out -o coverage-all.html

cmd/coordinator/cmd/cni_types.go

@@ -269,8 +269,8 @@ func validateRPFilterConfig(rpfilter *int32, coordinatorConfig int64) (*int32, e
 func ValidateDelectOptions(config *DetectOptions) (*DetectOptions, error) {
 	if config == nil {
 		return &DetectOptions{
-			Interval: "1s",
-			TimeOut:  "3s",
+			Interval: "10ms",
+			TimeOut:  "100ms",
 			Retry:    3,
 		}, nil
 	}
@@ -284,7 +284,7 @@ func ValidateDelectOptions(config *DetectOptions) (*DetectOptions, error) {
 	}
 
 	if config.TimeOut == "" {
-		config.TimeOut = "100ms"
+		config.TimeOut = "500ms"
 	}
 
 	_, err := time.ParseDuration(config.Interval)

docs/concepts/coordinator-zh_CN.md

@@ -40,7 +40,7 @@ Spiderpool 内置一个叫 `coordinator` 的 CNI meta-plugin, 它在 Main CNI 
 | hostRuleTable      | 策略路由表号，同主机与 Pod 通信的路由将会存放于这个表号                                                                                                                                                                                                                                          | 整数型      | optional   | 500                               |
 | podRPFilter       | 设置 Pod 的 sysctl 参数 rp_filter                                                                                                                                                                                                                                              | 整数型      | optional   | 0                                 |
 | hostRPFilter       | (遗弃)设置节点 的 sysctl 参数 rp_filter                                                                                                                                                                                                                                              | 整数型      | optional   | 0                                 |
-| detectOptions      | 检测地址冲突和网关可达性的高级配置项: 包括重试次数(默认为 3 次), 探测间隔(默认为 10ms) 和 超时时间(默认为 100ms)                                                                                                                                                                                                        | 对象类型     | optional   | 空                                 |
+| detectOptions      | 检测地址冲突和网关可达性的高级配置项: 包括发送探测报文次数(retries: 默认为 3 次), 和响应的超时时间(timeout: 默认为 100ms)，还有发送报文的间隔(interval:默认为 10ms, 将会在未来版本中移除)                                                                                                                                                                                                       | 对象类型     | optional   | 空                                 |
 | logOptions         | 日志配置，包括 logLevel(默认为 debug) 和 logFile(默认为 /var/log/spidernet/coordinator.log)                                                                                                                                                                                           | 对象类型     | optional   | -                                 |
 
 > 如果您通过 `SpinderMultusConfig CR`  帮助创建 NetworkAttachmentDefinition CR，您可以在 `SpinderMultusConfig` 中配置 `coordinator` (所有字段)。参考: [SpinderMultusConfig](../reference/crd-spidermultusconfig.md)。

docs/concepts/coordinator.md

@@ -41,7 +41,7 @@ Let's delve into how coordinator implements these features.
 | hostRuleTable      | The routes on the host that communicates with the pod's underlay IPs will belong to this routing table number                                                                                                                                                                                                                                                                                                                                                                                                                                                                           | int      | optional   | 500         |
 | podRPFilter       | Set the rp_filter sysctl parameter on the pod, which is recommended to be set to 0                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     | int      | optional   | 0           |
 | hostRPFilter      | (deprecated)Set the rp_filter sysctl parameter on the node, which is recommended to be set to 0                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     | int      | optional   | 0           |
-| detectOptions      | The advanced configuration of detectGateway and detectIPConflict, including retry numbers(default is 3), interval(default is 10ms) and timeout(default is 100ms)                                                                                                                                                                                                                                                                                                                                                                                                                             | obejct   | optional   | nil         |
+| detectOptions      | The advanced configuration of detectGateway and detectIPConflict, including the number of the send packets(retries: default is 3) and the response timeout(timeout: default is 100ms) and the packet sending interval(interval: default is 10ms, which will be removed in the future version).                                                                                                                                                                                                                                                                                                                                                                                                                            | obejct   | optional   | nil         |
 | logOptions         | The configuration of logging, including logLevel(default is debug) and logFile(default is /var/log/spidernet/coordinator.log)                                                                                                                                                                                                                                                                                                                                                                                                                                                           | obejct   | optional   | nil         |
 
 > You can configure `coordinator` by specifying all the relevant fields in `SpinderMultusConfig` if a NetworkAttachmentDefinition CR is created via `SpinderMultusConfig CR`. For more information, please refer to [SpinderMultusConfig](../reference/crd-spidermultusconfig.md).

pkg/coordinatormanager/coordinator_informer.go

@@ -8,7 +8,6 @@ import (
 	"fmt"
 	"net"
 	"reflect"
-	"regexp"
 	"sort"
 	"strings"
 	"time"
@@ -470,7 +469,7 @@ func (cc *CoordinatorController) updatePodAndServerCIDR(ctx context.Context, log
 
 	if err == nil {
 		logger.Sugar().Info("Trying to fetch the ClusterCIDR from kube-system/kubeadm-config")
-		k8sPodCIDR, k8sServiceCIDR, err = ExtractK8sCIDRFromKubeadmConfigMap(&cm)
+		k8sPodCIDR, k8sServiceCIDR, err = utils.ExtractK8sCIDRFromKubeadmConfigMap(&cm)
 		if err == nil {
 			// Success to get ClusterCIDR from kubeadm-config
 			logger.Sugar().Infof("Success get CIDR from kubeadm-config: PodCIDR=%v, ServiceCIDR=%v", k8sPodCIDR, k8sServiceCIDR)
@@ -504,7 +503,7 @@ func (cc *CoordinatorController) updatePodAndServerCIDR(ctx context.Context, log
 			return coordCopy
 		}
 
-		k8sPodCIDR, k8sServiceCIDR = ExtractK8sCIDRFromKCMPod(&podList.Items[0])
+		k8sPodCIDR, k8sServiceCIDR = utils.ExtractK8sCIDRFromKCMPod(&podList.Items[0])
 		logger.Sugar().Infof("kube-controller-manager k8sPodCIDR %v, k8sServiceCIDR %v", k8sPodCIDR, k8sServiceCIDR)
 	}
 
@@ -781,90 +780,6 @@ func (cc *CoordinatorController) updateServiceCIDR(logger *zap.Logger, coordCopy
 	return nil
 }
 
-func ExtractK8sCIDRFromKubeadmConfigMap(cm *corev1.ConfigMap) ([]string, []string, error) {
-	if cm == nil {
-		return nil, nil, fmt.Errorf("kubeadm configmap is unexpected to nil")
-	}
-	var podCIDR, serviceCIDR []string
-
-	clusterConfig, exists := cm.Data["ClusterConfiguration"]
-	if !exists {
-		return podCIDR, serviceCIDR, fmt.Errorf("unable to get kubeadm configmap ClusterConfiguration")
-	}
-
-	podReg := regexp.MustCompile(`podSubnet:\s*(\S+)`)
-	serviceReg := regexp.MustCompile(`serviceSubnet:\s*(\S+)`)
-
-	podSubnets := podReg.FindStringSubmatch(clusterConfig)
-	serviceSubnets := serviceReg.FindStringSubmatch(clusterConfig)
-
-	if len(podSubnets) > 1 {
-		for _, cidr := range strings.Split(podSubnets[1], ",") {
-			cidr = strings.TrimSpace(cidr)
-			_, _, err := net.ParseCIDR(cidr)
-			if err != nil {
-				continue
-			}
-			podCIDR = append(podCIDR, cidr)
-		}
-	}
-
-	if len(serviceSubnets) > 1 {
-		for _, cidr := range strings.Split(serviceSubnets[1], ",") {
-			cidr = strings.TrimSpace(cidr)
-			_, _, err := net.ParseCIDR(cidr)
-			if err != nil {
-				continue
-			}
-			serviceCIDR = append(serviceCIDR, cidr)
-		}
-	}
-
-	return podCIDR, serviceCIDR, nil
-}
-
-func ExtractK8sCIDRFromKCMPod(kcm *corev1.Pod) ([]string, []string) {
-	var podCIDR, serviceCIDR []string
-
-	podReg := regexp.MustCompile(`--cluster-cidr=(.*)`)
-	serviceReg := regexp.MustCompile(`--service-cluster-ip-range=(.*)`)
-
-	var podSubnets, serviceSubnets []string
-	for _, l := range kcm.Spec.Containers[0].Command {
-		if len(podSubnets) == 0 {
-			podSubnets = podReg.FindStringSubmatch(l)
-		}
-		if len(serviceSubnets) == 0 {
-			serviceSubnets = serviceReg.FindStringSubmatch(l)
-		}
-		if len(podSubnets) != 0 && len(serviceSubnets) != 0 {
-			break
-		}
-	}
-
-	if len(podSubnets) != 0 {
-		for _, cidr := range strings.Split(podSubnets[1], ",") {
-			_, _, err := net.ParseCIDR(cidr)
-			if err != nil {
-				continue
-			}
-			podCIDR = append(podCIDR, cidr)
-		}
-	}
-
-	if len(serviceSubnets) != 0 {
-		for _, cidr := range strings.Split(serviceSubnets[1], ",") {
-			_, _, err := net.ParseCIDR(cidr)
-			if err != nil {
-				continue
-			}
-			serviceCIDR = append(serviceCIDR, cidr)
-		}
-	}
-
-	return podCIDR, serviceCIDR
-}
-
 func fetchType(cniDir string) (string, error) {
 	defaultCniName, err := utils.GetDefaultCniName(cniDir)
 	if err != nil {

pkg/errgroup/errgroup.go

@@ -11,6 +11,7 @@
 package errgroup
 
 import (
+	"context"
 	"fmt"
 	"runtime"
 	"sync"
@@ -43,6 +44,16 @@ func (g *Group) done() {
 	g.wg.Done()
 }
 
+// WithContext returns a new Group and an associated Context derived from ctx.
+//
+// The derived Context is canceled the first time a function passed to Go
+// returns a non-nil error or the first time Wait returns, whichever occurs
+// first.
+func WithContext(ctx context.Context) (*Group, context.Context) {
+	ctx, cancel := context.WithCancelCause(ctx)
+	return &Group{cancel: cancel}, ctx
+}
+
 // Wait blocks until all function calls from the Go method have returned, then
 // returns the first non-nil error (if any) from them.
 func (g *Group) Wait() error {

pkg/ippoolmanager/config_test.go

@@ -0,0 +1,77 @@
+// Copyright 2024 Authors of spidernet-io
+// SPDX-License-Identifier: Apache-2.0
+package ippoolmanager
+
+import (
+	. "github.com/onsi/ginkgo/v2"
+	. "github.com/onsi/gomega"
+)
+
+var _ = Describe("IPPoolManagerConfig", Label("ippool_manager_test"), func() {
+	var config IPPoolManagerConfig
+
+	Describe("setDefaultsForIPPoolManagerConfig", func() {
+		Context("when MaxAllocatedIPs is nil", func() {
+			BeforeEach(func() {
+				config = IPPoolManagerConfig{
+					MaxAllocatedIPs:        nil,
+					EnableKubevirtStaticIP: false,
+				}
+			})
+
+			It("should set MaxAllocatedIPs to default value", func() {
+				result := setDefaultsForIPPoolManagerConfig(config)
+				Expect(result.MaxAllocatedIPs).NotTo(BeNil())
+				Expect(*result.MaxAllocatedIPs).To(Equal(defaultMaxAllocatedIPs))
+			})
+		})
+
+		Context("when MaxAllocatedIPs is set", func() {
+			BeforeEach(func() {
+				maxIPs := 3000
+				config = IPPoolManagerConfig{
+					MaxAllocatedIPs:        &maxIPs,
+					EnableKubevirtStaticIP: true,
+				}
+			})
+
+			It("should retain the provided MaxAllocatedIPs value", func() {
+				result := setDefaultsForIPPoolManagerConfig(config)
+				Expect(result.MaxAllocatedIPs).NotTo(BeNil())
+				Expect(*result.MaxAllocatedIPs).To(Equal(3000))
+			})
+		})
+
+		Context("when EnableKubevirtStaticIP is true", func() {
+			BeforeEach(func() {
+				config = IPPoolManagerConfig{
+					MaxAllocatedIPs:        nil,
+					EnableKubevirtStaticIP: true,
+				}
+			})
+
+			It("should set MaxAllocatedIPs to default value", func() {
+				result := setDefaultsForIPPoolManagerConfig(config)
+				Expect(result.MaxAllocatedIPs).NotTo(BeNil())
+				Expect(*result.MaxAllocatedIPs).To(Equal(defaultMaxAllocatedIPs))
+				Expect(result.EnableKubevirtStaticIP).To(BeTrue())
+			})
+		})
+
+		Context("when EnableKubevirtStaticIP is false", func() {
+			BeforeEach(func() {
+				config = IPPoolManagerConfig{
+					MaxAllocatedIPs:        nil,
+					EnableKubevirtStaticIP: false,
+				}
+			})
+
+			It("should set MaxAllocatedIPs to default value", func() {
+				result := setDefaultsForIPPoolManagerConfig(config)
+				Expect(result.MaxAllocatedIPs).NotTo(BeNil())
+				Expect(*result.MaxAllocatedIPs).To(Equal(defaultMaxAllocatedIPs))
+				Expect(result.EnableKubevirtStaticIP).To(BeFalse())
+			})
+		})
+	})
+})