[ML] call job validation endpoint with complete payload

elastic · Apr 2, 2020 · dd997c4 · dd997c4
1 parent a729b3b
commit dd997c4
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 3 deletions.
diff --git a/x-pack/plugins/ml/public/application/services/ml_api_service/index.ts b/x-pack/plugins/ml/public/application/services/ml_api_service/index.ts
@@ -151,8 +151,15 @@ export const ml = {
     });
   },
 
-  validateJob({ job }: { job: Job }) {
-    const body = JSON.stringify({ job });
+  validateJob(payload: {
+    job: Job;
+    duration: {
+      start?: number;
+      end?: number;
+    };
+    fields?: any[];
+  }) {
+    const body = JSON.stringify(payload);
     return http<any>({
       path: `${basePath()}/validate/job`,
       method: 'POST',

diff --git a/x-pack/plugins/ml/server/models/job_validation/validate_model_memory_limit.ts b/x-pack/plugins/ml/server/models/job_validation/validate_model_memory_limit.ts
@@ -46,7 +46,7 @@ export async function validateModelMemoryLimit(
 
   // if there is no duration, do not run the estimate test
   const runCalcModelMemoryTest =
-    duration && typeof duration?.start !== undefined && duration?.end !== undefined;
+    duration && duration?.start !== undefined && duration?.end !== undefined;
 
   // retrieve the max_model_memory_limit value from the server
   // this will be unset unless the user has set this on their cluster