[SO migration] fail the migration if unknown types are encountered (e…

…lastic#118300)

* manually revert elastic#105213

* enabled xpack to register types for some IT tests

* fiz doc size

Co-authored-by: Kibana Machine <42973632+kibanamachine@users.noreply.github.com>

src/core/server/saved_objects/migrations/actions/check_for_unknown_docs.test.ts

@@ -98,12 +98,10 @@ describe('checkForUnknownDocs', () => {
     const result = await task();
 
     expect(Either.isRight(result)).toBe(true);
-    expect((result as Either.Right<any>).right).toEqual({
-      unknownDocs: [],
-    });
+    expect((result as Either.Right<any>).right).toEqual({});
   });
 
-  it('resolves with `Either.right` when unknown docs are found', async () => {
+  it('resolves with `Either.left` when unknown docs are found', async () => {
     const client = elasticsearchClientMock.createInternalClient(
       elasticsearchClientMock.createSuccessTransportRequestPromise({
         hits: {
@@ -124,8 +122,9 @@ describe('checkForUnknownDocs', () => {
 
     const result = await task();
 
-    expect(Either.isRight(result)).toBe(true);
-    expect((result as Either.Right<any>).right).toEqual({
+    expect(Either.isLeft(result)).toBe(true);
+    expect((result as Either.Left<any>).left).toEqual({
+      type: 'unknown_docs_found',
       unknownDocs: [
         { id: '12', type: 'foo' },
         { id: '14', type: 'bar' },
@@ -151,8 +150,9 @@ describe('checkForUnknownDocs', () => {
 
     const result = await task();
 
-    expect(Either.isRight(result)).toBe(true);
-    expect((result as Either.Right<any>).right).toEqual({
+    expect(Either.isLeft(result)).toBe(true);
+    expect((result as Either.Left<any>).left).toEqual({
+      type: 'unknown_docs_found',
       unknownDocs: [{ id: '12', type: 'unknown' }],
     });
   });

src/core/server/saved_objects/migrations/actions/check_for_unknown_docs.ts

@@ -32,6 +32,7 @@ export interface CheckForUnknownDocsFoundDoc {
 
 /** @internal */
 export interface UnknownDocsFound {
+  type: 'unknown_docs_found';
   unknownDocs: CheckForUnknownDocsFoundDoc[];
 }
 
@@ -41,7 +42,10 @@ export const checkForUnknownDocs =
     indexName,
     unusedTypesQuery,
     knownTypes,
-  }: CheckForUnknownDocsParams): TaskEither.TaskEither<RetryableEsClientError, UnknownDocsFound> =>
+  }: CheckForUnknownDocsParams): TaskEither.TaskEither<
+    RetryableEsClientError | UnknownDocsFound,
+    {}
+  > =>
   () => {
     const query = createUnknownDocQuery(unusedTypesQuery, knownTypes);
 
@@ -54,9 +58,14 @@ export const checkForUnknownDocs =
       })
       .then((response) => {
         const { hits } = response.body.hits;
-        return Either.right({
-          unknownDocs: hits.map((hit) => ({ id: hit._id, type: hit._source?.type ?? 'unknown' })),
-        });
+        if (hits.length) {
+          return Either.left({
+            type: 'unknown_docs_found' as const,
+            unknownDocs: hits.map((hit) => ({ id: hit._id, type: hit._source?.type ?? 'unknown' })),
+          });
+        } else {
+          return Either.right({});
+        }
       })
       .catch(catchRetryableEsClientErrors);
   };

src/core/server/saved_objects/migrations/actions/index.ts

@@ -80,6 +80,7 @@ export type {
 } from './update_and_pickup_mappings';
 export { updateAndPickupMappings } from './update_and_pickup_mappings';
 
+import type { UnknownDocsFound } from './check_for_unknown_docs';
 export type {
   CheckForUnknownDocsParams,
   UnknownDocsFound,
@@ -141,6 +142,7 @@ export interface ActionErrorTypeMap {
   remove_index_not_a_concrete_index: RemoveIndexNotAConcreteIndex;
   documents_transform_failed: DocumentsTransformFailed;
   request_entity_too_large_exception: RequestEntityTooLargeException;
+  unknown_docs_found: UnknownDocsFound;
 }
 
 /**

src/core/server/saved_objects/migrations/core/migrate_raw_docs.test.ts

@@ -139,7 +139,6 @@ describe('migrateRawDocsSafely', () => {
     ]);
     const task = migrateRawDocsSafely({
       serializer: new SavedObjectsSerializer(new SavedObjectTypeRegistry()),
-      knownTypes: new Set(['a', 'c']),
       migrateDoc: transform,
       rawDocs: [
         { _id: 'a:b', _source: { type: 'a', a: { name: 'AAA' } } },
@@ -184,7 +183,6 @@ describe('migrateRawDocsSafely', () => {
     ]);
     const task = migrateRawDocsSafely({
       serializer: new SavedObjectsSerializer(new SavedObjectTypeRegistry()),
-      knownTypes: new Set(['a', 'c']),
       migrateDoc: transform,
       rawDocs: [
         { _id: 'foo:b', _source: { type: 'a', a: { name: 'AAA' } } },
@@ -206,7 +204,6 @@ describe('migrateRawDocsSafely', () => {
     ]);
     const task = migrateRawDocsSafely({
       serializer: new SavedObjectsSerializer(new SavedObjectTypeRegistry()),
-      knownTypes: new Set(['a', 'c']),
       migrateDoc: transform,
       rawDocs: [{ _id: 'a:b', _source: { type: 'a', a: { name: 'AAA' } } }],
     });
@@ -240,7 +237,6 @@ describe('migrateRawDocsSafely', () => {
     });
     const task = migrateRawDocsSafely({
       serializer: new SavedObjectsSerializer(new SavedObjectTypeRegistry()),
-      knownTypes: new Set(['a', 'c']),
       migrateDoc: transform,
       rawDocs: [{ _id: 'a:b', _source: { type: 'a', a: { name: 'AAA' } } }], // this is the raw doc
     });
@@ -256,43 +252,4 @@ describe('migrateRawDocsSafely', () => {
       }
     `);
   });
-
-  test('skips documents of unknown types', async () => {
-    const transform = jest.fn<any, any>((doc: any) => [
-      set(_.cloneDeep(doc), 'attributes.name', 'HOI!'),
-    ]);
-    const task = migrateRawDocsSafely({
-      serializer: new SavedObjectsSerializer(new SavedObjectTypeRegistry()),
-      knownTypes: new Set(['a']),
-      migrateDoc: transform,
-      rawDocs: [
-        { _id: 'a:b', _source: { type: 'a', a: { name: 'AAA' } } },
-        { _id: 'c:d', _source: { type: 'c', c: { name: 'DDD' } } },
-      ],
-    });
-
-    const result = (await task()) as Either.Right<DocumentsTransformSuccess>;
-    expect(result._tag).toEqual('Right');
-    expect(result.right.processedDocs).toEqual([
-      {
-        _id: 'a:b',
-        _source: { type: 'a', a: { name: 'HOI!' }, migrationVersion: {}, references: [] },
-      },
-      {
-        _id: 'c:d',
-        // name field is not migrated on unknown type
-        _source: { type: 'c', c: { name: 'DDD' } },
-      },
-    ]);
-
-    const obj1 = {
-      id: 'b',
-      type: 'a',
-      attributes: { name: 'AAA' },
-      migrationVersion: {},
-      references: [],
-    };
-    expect(transform).toHaveBeenCalledTimes(1);
-    expect(transform).toHaveBeenNthCalledWith(1, obj1);
-  });
 });

src/core/server/saved_objects/migrations/core/migrate_raw_docs.ts

@@ -25,13 +25,16 @@ export interface DocumentsTransformFailed {
   readonly corruptDocumentIds: string[];
   readonly transformErrors: TransformErrorObjects[];
 }
+
 export interface DocumentsTransformSuccess {
   readonly processedDocs: SavedObjectsRawDoc[];
 }
+
 export interface TransformErrorObjects {
   readonly rawId: string;
   readonly err: TransformSavedObjectDocumentError | Error;
 }
+
 type MigrateFn = (
   doc: SavedObjectUnsanitizedDoc<unknown>
 ) => Promise<Array<SavedObjectUnsanitizedDoc<unknown>>>;
@@ -83,7 +86,6 @@ export async function migrateRawDocs(
 
 interface MigrateRawDocsSafelyDeps {
   serializer: SavedObjectsSerializer;
-  knownTypes: ReadonlySet<string>;
   migrateDoc: MigrateAndConvertFn;
   rawDocs: SavedObjectsRawDoc[];
 }
@@ -97,7 +99,6 @@ interface MigrateRawDocsSafelyDeps {
  */
 export function migrateRawDocsSafely({
   serializer,
-  knownTypes,
   migrateDoc,
   rawDocs,
 }: MigrateRawDocsSafelyDeps): TaskEither.TaskEither<
@@ -111,10 +112,7 @@ export function migrateRawDocsSafely({
     const corruptSavedObjectIds: string[] = [];
     const options = { namespaceTreatment: 'lax' as const };
     for (const raw of rawDocs) {
-      // Do not transform documents of unknown types
-      if (raw?._source?.type && !knownTypes.has(raw._source.type)) {
-        processedDocs.push(raw);
-      } else if (serializer.isRawSavedObject(raw, options)) {
+      if (serializer.isRawSavedObject(raw, options)) {
         try {
           const savedObject = convertToRawAddMigrationVersion(raw, options, serializer);
           processedDocs.push(