Merge branch 'master' into check-before-pull

Author: bfournie

.ci-operator.yaml

@@ -1,4 +1,4 @@
 build_root_image:
   name: release
   namespace: openshift
-  tag: rhel-8-release-golang-1.19-openshift-4.12
+  tag: rhel-8-release-golang-1.19-openshift-4.13

cmd/openshift-install/agent.go

@@ -56,7 +56,7 @@ var (
 	}
 
 	agentImageTarget = target{
-		name: "Image",
+		name: "Agent ISO Image",
 		command: &cobra.Command{
 			Use:   "image",
 			Short: "Generates a bootable image containing all the information needed to deploy a cluster",
@@ -69,7 +69,21 @@ var (
 		},
 	}
 
-	agentTargets = []target{agentConfigTarget, agentManifestsTarget, agentImageTarget}
+	agentPXEFilesTarget = target{
+		name: "Agent PXE Files",
+		command: &cobra.Command{
+			Use:   "pxe-files",
+			Short: "Generates PXE bootable image files containing all the information needed to deploy a cluster",
+			Args:  cobra.ExactArgs(0),
+		},
+		assets: []asset.WritableAsset{
+			&image.AgentPXEFiles{},
+			&kubeconfig.AgentAdminClient{},
+			&password.KubeadminPassword{},
+		},
+	}
+
+	agentTargets = []target{agentConfigTarget, agentManifestsTarget, agentImageTarget, agentPXEFilesTarget}
 )
 
 func newAgentCreateCmd() *cobra.Command {

cmd/openshift-install/agent/waitfor.go

@@ -1,6 +1,8 @@
 package agent
 
 import (
+	"context"
+
 	"github.com/pkg/errors"
 	"github.com/sirupsen/logrus"
 	"github.com/spf13/cobra"
@@ -31,6 +33,19 @@ func NewWaitForCmd() *cobra.Command {
 	return cmd
 }
 
+func handleBootstrapError(cluster *agentpkg.Cluster, err error) {
+	logrus.Debug("Printing the event list gathered from the Agent Rest API")
+	cluster.PrintInfraEnvRestAPIEventList()
+	err2 := cluster.API.OpenShift.LogClusterOperatorConditions()
+	if err2 != nil {
+		logrus.Error("Attempted to gather ClusterOperator status after wait failure: ", err2)
+	}
+	logrus.Info("Use the following commands to gather logs from the cluster")
+	logrus.Info("openshift-install gather bootstrap --help")
+	logrus.Error(errors.Wrap(err, "Bootstrap failed to complete: "))
+	logrus.Exit(exitCodeBootstrapFailed)
+}
+
 func newWaitForBootstrapCompleteCmd() *cobra.Command {
 	return &cobra.Command{
 		Use:   "bootstrap-complete",
@@ -42,19 +57,16 @@ func newWaitForBootstrapCompleteCmd() *cobra.Command {
 			if len(assetDir) == 0 {
 				logrus.Fatal("No cluster installation directory found")
 			}
-			cluster, err := agentpkg.WaitForBootstrapComplete(assetDir)
+
+			ctx := context.Background()
+			cluster, err := agentpkg.NewCluster(ctx, assetDir)
 			if err != nil {
-				logrus.Debug("Printing the event list gathered from the Agent Rest API")
-				cluster.PrintInfraEnvRestAPIEventList()
-				err2 := cluster.API.OpenShift.LogClusterOperatorConditions()
-				if err2 != nil {
-					logrus.Error("Attempted to gather ClusterOperator status after wait failure: ", err2)
-				}
-				logrus.Info("Use the following commands to gather logs from the cluster")
-				logrus.Info("openshift-install gather bootstrap --help")
-				logrus.Error(errors.Wrap(err, "Bootstrap failed to complete: "))
 				logrus.Exit(exitCodeBootstrapFailed)
 			}
+
+			if err := agentpkg.WaitForBootstrapComplete(cluster); err != nil {
+				handleBootstrapError(cluster, err)
+			}
 		},
 	}
 }
@@ -70,10 +82,19 @@ func newWaitForInstallCompleteCmd() *cobra.Command {
 			if len(assetDir) == 0 {
 				logrus.Fatal("No cluster installation directory found")
 			}
-			cluster, err := agentpkg.WaitForInstallComplete(assetDir)
+
+			ctx := context.Background()
+			cluster, err := agentpkg.NewCluster(ctx, assetDir)
 			if err != nil {
-				logrus.Debug("Printing the event list gathered from the Agent Rest API")
-				cluster.PrintInfraEnvRestAPIEventList()
+				logrus.Exit(exitCodeBootstrapFailed)
+			}
+
+			if err := agentpkg.WaitForBootstrapComplete(cluster); err != nil {
+				handleBootstrapError(cluster, err)
+			}
+
+			if err = agentpkg.WaitForInstallComplete(cluster); err != nil {
+				logrus.Error(err)
 				err2 := cluster.API.OpenShift.LogClusterOperatorConditions()
 				if err2 != nil {
 					logrus.Error("Attempted to gather ClusterOperator status after wait failure: ", err2)

cmd/openshift-install/agent_internal_integration_test.go

@@ -103,11 +103,27 @@ func runIntegrationTest(t *testing.T, testFolder string) {
 		},
 
 		Cmds: map[string]func(*testscript.TestScript, bool, []string){
-			"isocmp": isoCmp,
+			"isocmp":              isoCmp,
+			"ignitionImgContains": ignitionImgContains,
 		},
 	})
 }
 
+// [!] ignitionImgContains `isoPath` `file` check if the specified file `file`
+// is stored within /images/ignition.img archive in the ISO `isoPath` image.
+func ignitionImgContains(ts *testscript.TestScript, neg bool, args []string) {
+	if len(args) != 2 {
+		ts.Fatalf("usage: ignitionImgContains isoPath file")
+	}
+
+	workDir := ts.Getenv("WORK")
+	isoPath, eFilePath := args[0], args[1]
+	isoPathAbs := filepath.Join(workDir, isoPath)
+
+	_, err := extractFileFromIgnitionImg(isoPathAbs, eFilePath)
+	ts.Check(err)
+}
+
 // [!] isoCmp `isoPath` `isoFile` `expectedFile` check that the content of the file
 // `isoFile` - extracted from the ISO embedded ignition configuration file referenced
 // by `isoPath` - matches the content of the local file `expectedFile`.
@@ -120,7 +136,7 @@ func isoCmp(ts *testscript.TestScript, neg bool, args []string) {
 	isoPath, aFilePath, eFilePath := args[0], args[1], args[2]
 	isoPathAbs := filepath.Join(workDir, isoPath)
 
-	aData, err := readFileFromISO(isoPathAbs, aFilePath)
+	aData, err := readFileFromISOIgnitionCfg(isoPathAbs, aFilePath)
 	ts.Check(err)
 
 	eFilePathAbs := filepath.Join(workDir, eFilePath)
@@ -152,8 +168,8 @@ func isoCmp(ts *testscript.TestScript, neg bool, args []string) {
 	ts.Fatalf("%s and %s differ", aFilePath, eFilePath)
 }
 
-func readFileFromISO(isoPath string, nodePath string) ([]byte, error) {
-	config, err := readIgnitionFromISO(isoPath)
+func readFileFromISOIgnitionCfg(isoPath string, nodePath string) ([]byte, error) {
+	config, err := extractIgnitionCfg(isoPath)
 	if err != nil {
 		return nil, err
 	}
@@ -171,7 +187,7 @@ func readFileFromISO(isoPath string, nodePath string) ([]byte, error) {
 	return nil, errors.NotFound(nodePath)
 }
 
-func readIgnitionFromISO(isoPath string) (*igntypes.Config, error) {
+func extractFileFromIgnitionImg(isoPath string, fileName string) ([]byte, error) {
 	disk, err := diskfs.OpenWithMode(isoPath, diskfs.ReadOnly)
 	if err != nil {
 		return nil, err
@@ -193,12 +209,31 @@ func readIgnitionFromISO(isoPath string) (*igntypes.Config, error) {
 	}
 
 	cpioReader := cpio.NewReader(gzipReader)
-	_, err = cpioReader.Next()
-	if err != nil {
-		return nil, err
+
+	for {
+		header, err := cpioReader.Next()
+		if err == io.EOF { //nolint:errorlint
+			// end of cpio archive
+			break
+		}
+		if err != nil {
+			return nil, err
+		}
+
+		if header.Name == fileName {
+			rawContent, err := io.ReadAll(cpioReader)
+			if err != nil {
+				return nil, err
+			}
+			return rawContent, nil
+		}
 	}
 
-	rawContent, err := io.ReadAll(cpioReader)
+	return nil, errors.NotFound(fmt.Sprintf("File %s not found within the /images/ignition.img archive", fileName))
+}
+
+func extractIgnitionCfg(isoPath string) (*igntypes.Config, error) {
+	rawContent, err := extractFileFromIgnitionImg(isoPath, "config.ign")
 	if err != nil {
 		return nil, err
 	}

cmd/openshift-install/create.go

@@ -27,6 +27,7 @@ import (
 	configclient "github.com/openshift/client-go/config/clientset/versioned"
 	routeclient "github.com/openshift/client-go/route/clientset/versioned"
 	"github.com/openshift/installer/pkg/asset"
+	"github.com/openshift/installer/pkg/asset/agent/agentconfig"
 	"github.com/openshift/installer/pkg/asset/cluster"
 	"github.com/openshift/installer/pkg/asset/installconfig"
 	"github.com/openshift/installer/pkg/asset/logging"
@@ -299,7 +300,7 @@ func runTargetCmd(targets ...asset.WritableAsset) func(cmd *cobra.Command, args
 			logrus.Fatal(err)
 		}
 		switch cmd.Name() {
-		case "cluster", "image":
+		case "cluster", "image", "pxe-files":
 		default:
 			logrus.Infof(logging.LogCreatedFiles(cmd.Name(), rootOpts.dir, targets))
 		}
@@ -376,6 +377,11 @@ func waitForBootstrapComplete(ctx context.Context, config *rest.Config) *cluster
 	silenceRemaining := logDownsample
 	previousErrorSuffix := ""
 	timer.StartTimer("API")
+
+	if assetStore, err := assetstore.NewStore(rootOpts.dir); err == nil {
+		checkIfAgentCommand(assetStore)
+	}
+
 	var lastErr error
 	wait.Until(func() {
 		version, err := discovery.ServerVersion()
@@ -475,6 +481,8 @@ func waitForInitializedCluster(ctx context.Context, config *rest.Config) error {
 				timeout = 60 * time.Minute
 			}
 		}
+
+		checkIfAgentCommand(assetStore)
 	}
 
 	untilTime := time.Now().Add(timeout)
@@ -638,3 +646,9 @@ The cluster should be accessible for troubleshooting as detailed in the document
 https://docs.openshift.com/container-platform/latest/support/troubleshooting/troubleshooting-installations.html
 The 'wait-for install-complete' subcommand can then be used to continue the installation`)
 }
+
+func checkIfAgentCommand(assetStore asset.Store) {
+	if agentConfig, err := assetStore.Load(&agentconfig.AgentConfig{}); err == nil && agentConfig != nil {
+		logrus.Warning("An agent configuration was detected but this command is not the agent wait-for command")
+	}
+}

cmd/openshift-install/destroy.go

@@ -55,6 +55,7 @@ func newDestroyClusterCmd() *cobra.Command {
 			if err != nil {
 				logrus.Fatal(err)
 			}
+			logrus.Infof("Uninstallation complete!")
 		},
 	}
 }

cmd/openshift-install/main.go

@@ -9,7 +9,7 @@ import (
 	"github.com/pkg/errors"
 	"github.com/sirupsen/logrus"
 	"github.com/spf13/cobra"
-	"golang.org/x/crypto/ssh/terminal"
+	terminal "golang.org/x/term"
 	"k8s.io/klog"
 	klogv2 "k8s.io/klog/v2"
 )

cmd/openshift-install/testdata/agent/image/assets/default_assets.txt

@@ -0,0 +1,41 @@
+# Verify that the most relevant assets are properly generated in the ISO
+
+exec openshift-install agent create image --dir $WORK
+
+exists $WORK/agent.x86_64.iso
+
+ignitionImgContains agent.x86_64.iso config.ign
+ignitionImgContains agent.x86_64.iso agent
+
+-- install-config.yaml --
+apiVersion: v1
+baseDomain: test.metalkube.org
+controlPlane: 
+  name: master
+  replicas: 1
+compute: 
+- name: worker
+  replicas: 0
+metadata:
+  namespace: cluster0
+  name: ostest
+networking:
+  clusterNetwork:
+  - cidr: 10.128.0.0/14 
+    hostPrefix: 23 
+  networkType: OVNKubernetes
+  machineNetwork:
+  - cidr: 192.168.111.0/24
+  serviceNetwork: 
+  - 172.30.0.0/16
+platform:
+  none: {}
+sshKey: ssh-rsa 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
+pullSecret: '{"auths": {"quay.io": {"auth": "c3VwZXItc2VjcmV0Cg=="}}}'
+
+-- agent-config.yaml --
+apiVersion: v1alpha1
+metadata:
+  name: ostest
+  namespace: cluster0
+rendezvousIP: 192.168.111.20

cmd/openshift-install/testdata/agent/image/configurations/vsphere_compact.txt

@@ -0,0 +1,46 @@
+# Verify a default configuration for the compact topology on vsphere
+
+exec openshift-install agent create image --dir $WORK
+
+stderr 'The rendezvous host IP \(node0 IP\) is 192.168.111.20'
+
+exists $WORK/agent.x86_64.iso
+exists $WORK/auth/kubeconfig
+exists $WORK/auth/kubeadmin-password
+
+-- install-config.yaml --
+apiVersion: v1
+baseDomain: test.metalkube.org
+controlPlane:
+  name: master
+  replicas: 3
+compute:
+- name: worker
+  replicas: 0
+metadata:
+  namespace: cluster0
+  name: ostest
+networking:
+  clusterNetwork:
+  - cidr: 10.128.0.0/14
+    hostPrefix: 23
+  networkType: OVNKubernetes
+  machineNetwork:
+  - cidr: 192.168.111.0/24
+  serviceNetwork:
+  - 172.30.0.0/16
+platform:
+    vsphere:
+      apiVips:
+        - 192.168.111.5
+      ingressVips:
+        - 192.168.111.4
+sshKey: ssh-rsa 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
+pullSecret: '{"auths": {"quay.io": {"auth": "c3VwZXItc2VjcmV0Cg=="}}}'
+
+-- agent-config.yaml --
+apiVersion: v1alpha1
+metadata:
+  name: ostest
+  namespace: cluster0
+rendezvousIP: 192.168.111.20