feat: add schemaOnly to the import section (#2234)

The import feature was previously importing everything, schema and data,
with this new feature we will be able to select only schema or schema and data.

The way is doing this it's using the `section` feature from `pg_dump` and `pg_restore`
commands, we select which section we want to dump/restore, for a full dump and restore
we use `pre-data`, `data` and `post-data` sections, while to import only the schema
the section `pre-data` and `post-data` is used.

Closes #2228

Signed-off-by: Armando Ruocco <armando.ruocco@enterprisedb.com>
Signed-off-by: Gabriele Bartolini <gabriele.bartolini@enterprisedb.com>
Signed-off-by: Jonathan Gonzalez V <jonathan.gonzalez@enterprisedb.com>
Co-authored-by: Gabriele Bartolini <gabriele.bartolini@enterprisedb.com>
Co-authored-by: Jonathan Gonzalez V <jonathan.gonzalez@enterprisedb.com>
(cherry picked from commit ad0fdb2)

api/v1/cluster_types.go

@@ -1064,6 +1064,11 @@ type Import struct {
 	// database right after is imported - to be used with extreme care
 	// (by default empty). Only available in microservice type.
 	PostImportApplicationSQL []string `json:"postImportApplicationSQL,omitempty"`
+
+	// When set to true, only the `pre-data` and `post-data` sections of
+	// `pg_restore` are invoked, avoiding data import. Default: `false`.
+	// +kubebuilder:default:=false
+	SchemaOnly bool `json:"schemaOnly,omitempty"`
 }
 
 // ImportSource describes the source for the logical snapshot

config/crd/bases/postgresql.cnpg.io_clusters.yaml

@@ -1283,6 +1283,12 @@ spec:
                             items:
                               type: string
                             type: array
+                          schemaOnly:
+                            default: false
+                            description: 'When set to true, only the `pre-data` and
+                              `post-data` sections of `pg_restore` are invoked, avoiding
+                              data import. Default: `false`.'
+                            type: boolean
                           source:
                             description: The source of the import
                             properties:

docs/src/api_reference.md

@@ -540,6 +540,7 @@ Name                     | Description
 `databases               ` | The databases to import                                                                                                                                                                       - *mandatory*  | []string                     
 `roles                   ` | The roles to import                                                                                                                                                                           | []string                     
 `postImportApplicationSQL` | List of SQL queries to be executed as a superuser in the application database right after is imported - to be used with extreme care (by default empty). Only available in microservice type. | []string                     
+`schemaOnly              ` | When set to true, only the `pre-data` and `post-data` sections of `pg_restore` are invoked, avoiding data import. Default: `false`.                                                           | bool                         
 
 <a id='ImportSource'></a>
 

docs/src/samples/cluster-import-schema-only-basicauth.yaml

@@ -0,0 +1,27 @@
+apiVersion: postgresql.cnpg.io/v1
+kind: Cluster
+metadata:
+  name: cluster-pgdump
+spec:
+  instances: 3
+
+  bootstrap:
+    initdb:
+      import:
+        type: microservice
+        schemaOnly: true
+        databases:
+          - app
+        source:
+          externalCluster: cluster-example
+  storage:
+    size: 1Gi
+  externalClusters:
+    - name: cluster-example
+      connectionParameters:
+        host: cluster-example-rw.default.svc
+        user: postgres
+        dbname: postgres
+      password:
+        name: cluster-example-superuser
+        key: password

pkg/management/postgres/logicalimport/database.go

@@ -90,6 +90,12 @@ func (ds *databaseSnapshotter) exportDatabases(
 	databases []string,
 ) error {
 	contextLogger := log.FromContext(ctx)
+	sectionsToExport := []string{}
+
+	for _, section := range ds.getSectionsToExecute() {
+		sectionsToExport = append(sectionsToExport, fmt.Sprintf("--section=%s", section))
+	}
+
 	for _, database := range databases {
 		contextLogger.Info("exporting database", "databaseName", database)
 		dsn := target.GetDsn(database)
@@ -99,6 +105,7 @@ func (ds *databaseSnapshotter) exportDatabases(
 			"-d", dsn,
 			"-v",
 		}
+		options = append(options, sectionsToExport...)
 
 		contextLogger.Info("Running pg_dump", "cmd", pgDump,
 			"options", options)
@@ -120,7 +127,7 @@ func (ds *databaseSnapshotter) importDatabases(
 	contextLogger := log.FromContext(ctx)
 
 	for _, database := range databases {
-		for _, section := range []string{"pre-data", "data", "post-data"} {
+		for _, section := range ds.getSectionsToExecute() {
 			targetDatabase := target.GetDsn(database)
 			contextLogger.Info(
 				"executing database importing section",
@@ -190,7 +197,7 @@ func (ds *databaseSnapshotter) importDatabaseContent(
 		return err
 	}
 
-	for _, section := range []string{"pre-data", "data", "post-data"} {
+	for _, section := range ds.getSectionsToExecute() {
 		contextLogger.Info(
 			"executing database importing section",
 			"databaseName", database,
@@ -340,3 +347,20 @@ func (ds *databaseSnapshotter) dropExtensionsFromDatabase(
 
 	return rows.Err()
 }
+
+// getSectionsToExecute determines which stages of `pg_restore` and `pg_dump` to execute,
+// based on the configuration of the cluster. It returns a slice of strings representing
+// the sections to execute. These sections are labeled as "pre-data", "data", and "post-data".
+func (ds *databaseSnapshotter) getSectionsToExecute() []string {
+	const (
+		preData  = "pre-data"
+		data     = "data"
+		postData = "post-data"
+	)
+
+	if ds.cluster.Spec.Bootstrap.InitDB.Import.SchemaOnly {
+		return []string{preData, postData}
+	}
+
+	return []string{preData, data, postData}
+}