[8.19] [APM][ML] Limit environment name length when creating ML job (#225973)

x-pack/solutions/observability/plugins/apm/ftr_e2e/cypress/e2e/settings/anomaly_detection.cy.ts

@@ -75,6 +75,10 @@ const getAbleToModifyCase = () => {
   });
 
   it('should show error if api call crashes when modifying settings', () => {
+    cy.intercept('POST', '/internal/apm/settings/anomaly-detection/jobs', {
+      statusCode: 500,
+    });
+
     const { rangeFrom, rangeTo } = timeRange;
     const TEST_ENV =
       'Synthtrace: case scenario TEST-with-a-really-long-name ' + new Date().toISOString();

x-pack/solutions/observability/plugins/apm/server/lib/anomaly_detection/create_anomaly_detection_jobs.ts

@@ -23,6 +23,7 @@ import { APM_ML_JOB_GROUP, ML_MODULE_ID_APM_TRANSACTION } from './constants';
 import { getAnomalyDetectionJobs } from './get_anomaly_detection_jobs';
 
 const DEFAULT_TIMEOUT = '60s';
+const ENV_MAX_LENGTH = 40;
 
 export async function createAnomalyDetectionJobs({
   mlClient,
@@ -107,10 +108,11 @@ async function createAnomalyDetectionJob({
 }) {
   return withApmSpan('create_anomaly_detection_job', async () => {
     const randomToken = uuidv4().substr(-4);
+    const sanitizedEnvironment = snakeCase(environment).slice(0, ENV_MAX_LENGTH); // limit env name due to ML job ID length constraints (up to 64 chars in total)
 
     const anomalyDetectionJob = mlClient.modules.setup({
       moduleId: ML_MODULE_ID_APM_TRANSACTION,
-      prefix: `${APM_ML_JOB_GROUP}-${snakeCase(environment)}-${randomToken}-`,
+      prefix: `${APM_ML_JOB_GROUP}-${sanitizedEnvironment}-${randomToken}-`,
       groups: [APM_ML_JOB_GROUP],
       indexPatternName: apmMetricIndex,
       applyToAllSpaces: true,

x-pack/solutions/observability/test/apm_api_integration/tests/settings/anomaly_detection/write_user.spec.ts

@@ -58,6 +58,12 @@ export default function apiTest({ getService }: FtrProviderContext) {
         });
 
         describe('when calling create endpoint', () => {
+          beforeEach(async () => {
+            const res = await getJobs({ user });
+            const jobIds = res.body.jobs.map((job: any) => job.jobId);
+            await deleteJobs(jobIds);
+          });
+
           it('creates two jobs', async () => {
             await createJobs(['production', 'staging'], { user });
 
@@ -69,19 +75,22 @@ export default function apiTest({ getService }: FtrProviderContext) {
             });
           });
 
-          describe('with existing ML jobs', () => {
-            before(async () => {
-              await createJobs(['production', 'staging'], { user });
-            });
-            it('skips duplicate job creation', async () => {
-              await createJobs(['production', 'test'], { user });
+          it('creates job with long environment name', async () => {
+            const longEnvironmentName =
+              'Production: This Is A Deliberately Long Environment Name To Test Limits - 1234567890';
+            const { status } = await createJobs([longEnvironmentName], { user });
+            expect(status).to.be(200);
+          });
+
+          it('skips duplicate job creation', async () => {
+            await createJobs(['production', 'staging'], { user });
+            await createJobs(['production', 'test'], { user });
 
-              const { body } = await getJobs({ user });
-              expect(countBy(body.jobs, 'environment')).to.eql({
-                production: 1,
-                staging: 1,
-                test: 1,
-              });
+            const { body } = await getJobs({ user });
+            expect(countBy(body.jobs, 'environment')).to.eql({
+              production: 1,
+              staging: 1,
+              test: 1,
             });
           });
         });