fix: clean completed migration job before reinitialize on-prem demo (#…

…5890)

cmd/api-server/main.go

@@ -122,7 +122,7 @@ func runMigrations() (err error) {
 	return migrations.Migrator.Run(version.Version, migrator.MigrationTypeServer)
 }
 
-func runMongoMigrations(ctx context.Context, db *mongo.Database, migrationsDir string) error {
+func runMongoMigrations(ctx context.Context, db *mongo.Database, _ string) error {
 	migrationsCollectionName := "__migrations"
 	activeMigrations, err := dbmigrator.GetDbMigrationsFromFs(dbmigrations.MongoMigrationsFs)
 	if err != nil {

cmd/kubectl-testkube/commands/common/errors.go

@@ -34,6 +34,9 @@ const (
 	TKErrHelmCommandFailed ErrorCode = "TKERR-1301"
 	// TKErrKubectlCommandFailed is returned when a kubectl command fail.
 	TKErrKubectlCommandFailed ErrorCode = "TKERR-1302"
+
+	// TKErrCleanOldMigrationJobFailed is returned in case of issues with old migration jobs.
+	TKErrCleanOldMigrationJobFailed ErrorCode = "TKERR-1401"
 )
 
 const helpUrl = "https://testkubeworkspace.slack.com"

cmd/kubectl-testkube/commands/common/helper.go

@@ -162,6 +162,30 @@ func updateHelmRepo(helmPath string, dryRun bool, isOnPrem bool) *CLIError {
 	return nil
 }
 
+// It cleans existing migrations job with long TTL
+func CleanExistingCompletedMigrationJobs(namespace string) (cliErr *CLIError) {
+	kubectlPath, cliErr := lookupKubectlPath()
+	if cliErr != nil {
+		return cliErr
+	}
+
+	// Clean the job only when it's found and it's state is successful - ignore pending migrations.
+	succeeded, _ := runKubectlCommand(kubectlPath, []string{"get", "job", "testkube-enterprise-api-migrations", "-n", namespace, "-o", "jsonpath={.status.succeeded}"})
+	if succeeded == "1" {
+		_, err := runKubectlCommand(kubectlPath, []string{"delete", "job", "testkube-enterprise-api-migrations", "--namespace", namespace})
+		if err != nil {
+			return NewCLIError(
+				TKErrCleanOldMigrationJobFailed,
+				"Can't clean old migrations job",
+				"Migration job can't be deleted from some reason, check for errors in installation namespace, check execution. As a workaround try to delete job manually and retry installation/upgrade process",
+				err,
+			)
+		}
+	}
+
+	return nil
+}
+
 func runHelmCommand(helmPath string, args []string, dryRun bool) (commandOutput string, cliErr *CLIError) {
 	output, err := process.ExecuteWithOptions(process.Options{Command: helmPath, Args: args, DryRun: dryRun})
 	if err != nil {

cmd/kubectl-testkube/commands/init.go

@@ -214,6 +214,17 @@ func NewInitCmdDemo() *cobra.Command {
 				DemoValuesURL: demoValuesUrl,
 				DryRun:        dryRun,
 			}
+
+			cliErr = common.CleanExistingCompletedMigrationJobs(options.Namespace)
+			if cliErr != nil {
+				spinner.Fail("Failed to install Testkube On-Prem Demo")
+				if cfg.TelemetryEnabled {
+					cliErr.AddTelemetry(cmd, "installing", "install_failed", license)
+					_, _ = telemetry.HandleCLIErrorTelemetry(common.Version, cliErr)
+				}
+				common.HandleCLIError(cliErr)
+			}
+
 			cliErr = common.HelmUpgradeOrInstallTestkubeOnPremDemo(options)
 			if cliErr != nil {
 				spinner.Fail("Failed to install Testkube On-Prem Demo")

cmd/testworkflow-init/obfuscator/obfuscator.go

@@ -135,7 +135,7 @@ func (s *obfuscator) Write(p []byte) (n int, err error) {
 
 	// Write the rest of data
 	if len(p) > 0 {
-		nn, err = s.dst.Write(p)
+		_, err = s.dst.Write(p)
 		return size, err
 	}
 	return size, nil

pkg/agent/agent.go

@@ -378,7 +378,7 @@ func (ag *Agent) runWorkers(ctx context.Context, numWorkers int) error {
 	return g.Wait()
 }
 
-func (ag *Agent) executeCommand(ctx context.Context, cmd *cloud.ExecuteRequest) *cloud.ExecuteResponse {
+func (ag *Agent) executeCommand(_ context.Context, cmd *cloud.ExecuteRequest) *cloud.ExecuteResponse {
 	switch {
 	case cmd.Url == healthcheckCommand:
 		return &cloud.ExecuteResponse{MessageId: cmd.MessageId, Status: 0}