[7.x] [ML] Update cloning for jobs to use exclude_generated (elastic#…

…88898) (elastic#90014)

x-pack/plugins/ml/common/types/anomaly_detection_jobs/combined_job.ts

@@ -4,7 +4,6 @@
  * you may not use this file except in compliance with the Elastic License.
  */
 
-import { cloneDeep } from 'lodash';
 import { Datafeed } from './datafeed';
 import { DatafeedStats } from './datafeed_stats';
 import { Job } from './job';
@@ -25,16 +24,6 @@ export interface CombinedJobWithStats extends JobWithStats {
   datafeed_config: DatafeedWithStats;
 }
 
-export function expandCombinedJobConfig(combinedJob: CombinedJob) {
-  const combinedJobClone = cloneDeep(combinedJob);
-  const job = combinedJobClone;
-  const datafeed = combinedJobClone.datafeed_config;
-  // @ts-expect-error
-  delete job.datafeed_config;
-
-  return { job, datafeed };
-}
-
 export function isCombinedJobWithStats(arg: any): arg is CombinedJobWithStats {
   return typeof arg.job_id === 'string';
 }

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_creation/page.tsx

@@ -62,7 +62,7 @@ export const Page: FC<Props> = ({ jobId }) => {
     if (currentIndexPattern) {
       (async function () {
         if (jobId !== undefined) {
-          const analyticsConfigs = await ml.dataFrameAnalytics.getDataFrameAnalytics(jobId);
+          const analyticsConfigs = await ml.dataFrameAnalytics.getDataFrameAnalytics(jobId, true);
           if (
             Array.isArray(analyticsConfigs.data_frame_analytics) &&
             analyticsConfigs.data_frame_analytics.length > 0

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_management/components/action_clone/clone_action_name.tsx

@@ -310,9 +310,6 @@ export type CloneDataFrameAnalyticsConfig = Omit<
  */
 export function extractCloningConfig({
   id,
-  version,
-  // eslint-disable-next-line @typescript-eslint/naming-convention
-  create_time,
   ...configToClone
 }: DeepReadonly<DataFrameAnalyticsConfig>): CloneDataFrameAnalyticsConfig {
   return (cloneDeep({

x-pack/plugins/ml/public/application/jobs/jobs_list/components/utils.js

@@ -36,6 +36,23 @@ export function loadFullJob(jobId) {
   });
 }
 
+export function loadJobForCloning(jobId) {
+  return new Promise((resolve, reject) => {
+    ml.jobs
+      .jobForCloning(jobId)
+      .then((resp) => {
+        if (resp) {
+          resolve(resp);
+        } else {
+          throw new Error(`Could not find job ${jobId}`);
+        }
+      })
+      .catch((error) => {
+        reject(error);
+      });
+  });
+}
+
 export function isStartable(jobs) {
   return jobs.some(
     (j) => j.datafeedState === DATAFEED_STATE.STOPPED && j.jobState !== JOB_STATE.CLOSING
@@ -180,44 +197,56 @@ function showResults(resp, action) {
 
 export async function cloneJob(jobId) {
   try {
-    const job = await loadFullJob(jobId);
-    if (job.custom_settings && job.custom_settings.created_by) {
+    const [{ job: cloneableJob, datafeed }, originalJob] = await Promise.all([
+      loadJobForCloning(jobId),
+      loadFullJob(jobId, false),
+    ]);
+    if (cloneableJob !== undefined && originalJob?.custom_settings?.created_by !== undefined) {
       // if the job is from a wizards, i.e. contains a created_by property
       // use tempJobCloningObjects to temporarily store the job
-      mlJobService.tempJobCloningObjects.job = job;
+      mlJobService.tempJobCloningObjects.createdBy = originalJob?.custom_settings?.created_by;
+      mlJobService.tempJobCloningObjects.job = cloneableJob;
 
       if (
-        job.data_counts.earliest_record_timestamp !== undefined &&
-        job.data_counts.latest_record_timestamp !== undefined &&
-        job.data_counts.latest_bucket_timestamp !== undefined
+        originalJob.data_counts.earliest_record_timestamp !== undefined &&
+        originalJob.data_counts.latest_record_timestamp !== undefined &&
+        originalJob.data_counts.latest_bucket_timestamp !== undefined
       ) {
         // if the job has run before, use the earliest and latest record timestamp
         // as the cloned job's time range
-        let start = job.data_counts.earliest_record_timestamp;
-        let end = job.data_counts.latest_record_timestamp;
+        let start = originalJob.data_counts.earliest_record_timestamp;
+        let end = originalJob.data_counts.latest_record_timestamp;
 
-        if (job.datafeed_config.aggregations !== undefined) {
+        if (originalJob.datafeed_config.aggregations !== undefined) {
           // if the datafeed uses aggregations the earliest and latest record timestamps may not be the same
           // as the start and end of the data in the index.
-          const bucketSpanMs = parseInterval(job.analysis_config.bucket_span).asMilliseconds();
+          const bucketSpanMs = parseInterval(
+            originalJob.analysis_config.bucket_span
+          ).asMilliseconds();
           // round down to the start of the nearest bucket
           start =
-            Math.floor(job.data_counts.earliest_record_timestamp / bucketSpanMs) * bucketSpanMs;
+            Math.floor(originalJob.data_counts.earliest_record_timestamp / bucketSpanMs) *
+            bucketSpanMs;
           // use latest_bucket_timestamp and add two bucket spans minus one ms
-          end = job.data_counts.latest_bucket_timestamp + bucketSpanMs * 2 - 1;
+          end = originalJob.data_counts.latest_bucket_timestamp + bucketSpanMs * 2 - 1;
         }
 
         mlJobService.tempJobCloningObjects.start = start;
         mlJobService.tempJobCloningObjects.end = end;
       }
     } else {
       // otherwise use the tempJobCloningObjects
-      mlJobService.tempJobCloningObjects.job = job;
+      mlJobService.tempJobCloningObjects.job = cloneableJob;
+      // resets the createdBy field in case it still retains previous settings
+      mlJobService.tempJobCloningObjects.createdBy = undefined;
+    }
+    if (datafeed !== undefined) {
+      mlJobService.tempJobCloningObjects.datafeed = datafeed;
     }
 
-    if (job.calendars) {
+    if (originalJob.calendars) {
       mlJobService.tempJobCloningObjects.calendars = await mlCalendarService.fetchCalendarsByIds(
-        job.calendars
+        originalJob.calendars
       );
     }
 

x-pack/plugins/ml/public/application/jobs/new_job/pages/index_or_search/preconfigured_job_redirect.ts

@@ -8,19 +8,19 @@ import { ApplicationStart } from 'kibana/public';
 import { IndexPatternsContract } from '../../../../../../../../../src/plugins/data/public';
 import { mlJobService } from '../../../../services/job_service';
 import { loadIndexPatterns, getIndexPatternIdFromName } from '../../../../util/index_utils';
-import { CombinedJob } from '../../../../../../common/types/anomaly_detection_jobs';
+import { Datafeed, Job } from '../../../../../../common/types/anomaly_detection_jobs';
 import { CREATED_BY_LABEL, JOB_TYPE } from '../../../../../../common/constants/new_job';
 
 export async function preConfiguredJobRedirect(
   indexPatterns: IndexPatternsContract,
   basePath: string,
   navigateToUrl: ApplicationStart['navigateToUrl']
 ) {
-  const { job } = mlJobService.tempJobCloningObjects;
-  if (job) {
+  const { createdBy, job, datafeed } = mlJobService.tempJobCloningObjects;
+  if (job && datafeed) {
     try {
       await loadIndexPatterns(indexPatterns);
-      const redirectUrl = getWizardUrlFromCloningJob(job);
+      const redirectUrl = getWizardUrlFromCloningJob(createdBy, job, datafeed);
       await navigateToUrl(`${basePath}/app/ml/${redirectUrl}`);
       return Promise.reject();
     } catch (error) {
@@ -33,8 +33,8 @@ export async function preConfiguredJobRedirect(
   }
 }
 
-function getWizardUrlFromCloningJob(job: CombinedJob) {
-  const created = job?.custom_settings?.created_by;
+function getWizardUrlFromCloningJob(createdBy: string | undefined, job: Job, datafeed: Datafeed) {
+  const created = createdBy;
   let page = '';
 
   switch (created) {
@@ -55,7 +55,7 @@ function getWizardUrlFromCloningJob(job: CombinedJob) {
       break;
   }
 
-  const indexPatternId = getIndexPatternIdFromName(job.datafeed_config.indices.join());
+  const indexPatternId = getIndexPatternIdFromName(datafeed.indices.join());
 
   return `jobs/new_job/${page}?index=${indexPatternId}&_g=()`;
 }

x-pack/plugins/ml/public/application/jobs/new_job/pages/new_job/page.tsx

@@ -37,7 +37,6 @@ import { useMlContext } from '../../../../contexts/ml';
 import { getTimeFilterRange } from '../../../../components/full_time_range_selector';
 import { getTimeBucketsFromCache } from '../../../../util/time_buckets';
 import { ExistingJobsAndGroups, mlJobService } from '../../../../services/job_service';
-import { expandCombinedJobConfig } from '../../../../../../common/types/anomaly_detection_jobs';
 import { newJobCapsService } from '../../../../services/new_job_capabilities_service';
 import { EVENT_RATE_FIELD_ID } from '../../../../../../common/types/fields';
 import { getNewJobDefaults } from '../../../../services/ml_server_info';
@@ -74,10 +73,11 @@ export const Page: FC<PageProps> = ({ existingJobsAndGroups, jobType }) => {
 
   if (mlJobService.tempJobCloningObjects.job !== undefined) {
     // cloning a job
-    const clonedJob = mlJobService.cloneJob(mlJobService.tempJobCloningObjects.job);
-    const { job, datafeed } = expandCombinedJobConfig(clonedJob);
+    const clonedJob = mlJobService.tempJobCloningObjects.job;
+    const clonedDatafeed = mlJobService.cloneDatafeed(mlJobService.tempJobCloningObjects.datafeed);
+
     initCategorizationSettings();
-    jobCreator.cloneFromExistingJob(job, datafeed);
+    jobCreator.cloneFromExistingJob(clonedJob, clonedDatafeed);
 
     // if we're not skipping the time range, this is a standard job clone, so wipe the jobId
     if (mlJobService.tempJobCloningObjects.skipTimeRangeStep === false) {

x-pack/plugins/ml/public/application/services/job_service.d.ts

@@ -6,7 +6,7 @@
 
 import { SearchResponse } from 'elasticsearch';
 import { TimeRange } from 'src/plugins/data/common/query/timefilter/types';
-import { CombinedJob } from '../../../common/types/anomaly_detection_jobs';
+import { CombinedJob, Datafeed } from '../../../common/types/anomaly_detection_jobs';
 import { Calendar } from '../../../common/types/calendars';
 
 export interface ExistingJobsAndGroups {
@@ -18,6 +18,8 @@ declare interface JobService {
   jobs: CombinedJob[];
   createResultsUrlForJobs: (jobs: any[], target: string, timeRange?: TimeRange) => string;
   tempJobCloningObjects: {
+    createdBy?: string;
+    datafeed?: Datafeed;
     job: any;
     skipTimeRangeStep: boolean;
     start?: number;
@@ -26,7 +28,7 @@ declare interface JobService {
   };
   skipTimeRangeStep: boolean;
   saveNewJob(job: any): Promise<any>;
-  cloneJob(job: any): any;
+  cloneDatafeed(datafeed: any): Datafeed;
   openJob(jobId: string): Promise<any>;
   saveNewDatafeed(datafeedConfig: any, jobId: string): Promise<any>;
   startDatafeed(

x-pack/plugins/ml/public/application/services/job_service.js

@@ -28,6 +28,8 @@ class JobService {
     // if populated when loading the job management page, the start datafeed modal
     // is automatically opened.
     this.tempJobCloningObjects = {
+      createdBy: undefined,
+      datafeed: undefined,
       job: undefined,
       skipTimeRangeStep: false,
       start: undefined,
@@ -325,67 +327,15 @@ class JobService {
     return ml.addJob({ jobId: job.job_id, job }).then(func).catch(func);
   }
 
-  cloneJob(job) {
-    // create a deep copy of a job object
-    // also remove items from the job which are set by the server and not needed
-    // in the future this formatting could be optional
-    const tempJob = cloneDeep(job);
-
-    // remove all of the items which should not be copied
-    // such as counts, state and times
-    delete tempJob.state;
-    delete tempJob.job_version;
-    delete tempJob.data_counts;
-    delete tempJob.create_time;
-    delete tempJob.finished_time;
-    delete tempJob.last_data_time;
-    delete tempJob.model_size_stats;
-    delete tempJob.node;
-    delete tempJob.average_bucket_processing_time_ms;
-    delete tempJob.model_snapshot_id;
-    delete tempJob.open_time;
-    delete tempJob.established_model_memory;
-    delete tempJob.calendars;
-    delete tempJob.timing_stats;
-    delete tempJob.forecasts_stats;
-    delete tempJob.assignment_explanation;
-
-    delete tempJob.analysis_config.use_per_partition_normalization;
-
-    each(tempJob.analysis_config.detectors, (d) => {
-      delete d.detector_index;
-    });
+  cloneDatafeed(datafeed) {
+    const tempDatafeed = cloneDeep(datafeed);
 
     // remove parts of the datafeed config which should not be copied
-    if (tempJob.datafeed_config) {
-      delete tempJob.datafeed_config.datafeed_id;
-      delete tempJob.datafeed_config.job_id;
-      delete tempJob.datafeed_config.state;
-      delete tempJob.datafeed_config.node;
-      delete tempJob.datafeed_config.timing_stats;
-      delete tempJob.datafeed_config.assignment_explanation;
-
-      // remove query_delay if it's between 60s and 120s
-      // the back-end produces a random value between 60 and 120 and so
-      // by deleting it, the back-end will produce a new random value
-      if (tempJob.datafeed_config.query_delay) {
-        const interval = parseInterval(tempJob.datafeed_config.query_delay);
-        if (interval !== null) {
-          const queryDelay = interval.asSeconds();
-          if (queryDelay > 60 && queryDelay < 120) {
-            delete tempJob.datafeed_config.query_delay;
-          }
-        }
-      }
+    if (tempDatafeed) {
+      delete tempDatafeed.datafeed_id;
+      delete tempDatafeed.job_id;
     }
-
-    // when jumping from a wizard to the advanced job creation,
-    // the wizard's created_by information should be stripped.
-    if (tempJob.custom_settings && tempJob.custom_settings.created_by) {
-      delete tempJob.custom_settings.created_by;
-    }
-
-    return tempJob;
+    return tempDatafeed;
   }
 
   // find a job based on the id

x-pack/plugins/ml/public/application/services/ml_api_service/data_frame_analytics.ts

@@ -52,11 +52,12 @@ interface JobsExistsResponse {
 }
 
 export const dataFrameAnalytics = {
-  getDataFrameAnalytics(analyticsId?: string) {
+  getDataFrameAnalytics(analyticsId?: string, excludeGenerated?: boolean) {
     const analyticsIdString = analyticsId !== undefined ? `/${analyticsId}` : '';
     return http<GetDataFrameAnalyticsResponse>({
       path: `${basePath()}/data_frame/analytics${analyticsIdString}`,
       method: 'GET',
+      ...(excludeGenerated ? { query: { excludeGenerated } } : {}),
     });
   },
   getDataFrameAnalyticsStats(analyticsId?: string) {

x-pack/plugins/ml/public/application/services/ml_api_service/jobs.ts

@@ -13,6 +13,8 @@ import {
   MlJobWithTimeRange,
   MlSummaryJobs,
   CombinedJobWithStats,
+  Job,
+  Datafeed,
 } from '../../../../common/types/anomaly_detection_jobs';
 import { JobMessage } from '../../../../common/types/audit_message';
 import { AggFieldNamePair } from '../../../../common/types/fields';
@@ -48,6 +50,15 @@ export const jobsApiProvider = (httpService: HttpService) => ({
     });
   },
 
+  jobForCloning(jobId: string) {
+    const body = JSON.stringify({ jobId });
+    return httpService.http<{ job?: Job; datafeed?: Datafeed } | undefined>({
+      path: `${basePath()}/jobs/job_for_cloning`,
+      method: 'POST',
+      body,
+    });
+  },
+
   jobs(jobIds: string[]) {
     const body = JSON.stringify({ jobIds });
     return httpService.http<CombinedJobWithStats[]>({