From b0539ed267fe6b25a08f8cfd0ee9bda75c16347c Mon Sep 17 00:00:00 2001
From: Dimitris Athanasiou <dimitris@elastic.co>
Date: Mon, 14 Dec 2020 20:55:24 +0200
Subject: [PATCH] [7.x][ML] DFA _explain API should not fail when none field is
 included (#66281)

This commit fixes an issue with DFA _explain API where if it is called
and no field is included, it results to an error message coming from
the c++ process due to the data frame having no columns.

We want the _explain API not to error when no fields are included
exactly in order to explain to the user why it is that no fields
are included. Thus, we can simply fix this by not running the
memory estimation process and returning zero estimates instead.

Note that the _start API will fail with a user friendly error message
that informs there are no included fields.

Backport of #66281
---
 ...nsportExplainDataFrameAnalyticsAction.java |  8 ++++-
 .../test/ml/explain_data_frame_analytics.yml  | 32 +++++++++++++++++++
 2 files changed, 39 insertions(+), 1 deletion(-)

diff --git a/x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/action/TransportExplainDataFrameAnalyticsAction.java b/x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/action/TransportExplainDataFrameAnalyticsAction.java
index c0ccd42d38af6..a982bed0d7495 100644
--- a/x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/action/TransportExplainDataFrameAnalyticsAction.java
+++ b/x-pack/plugin/ml/src/main/java/org/elasticsearch/xpack/ml/action/TransportExplainDataFrameAnalyticsAction.java
@@ -17,6 +17,7 @@
 import org.elasticsearch.common.collect.Tuple;
 import org.elasticsearch.common.inject.Inject;
 import org.elasticsearch.common.settings.Settings;
+import org.elasticsearch.common.unit.ByteSizeValue;
 import org.elasticsearch.license.LicenseUtils;
 import org.elasticsearch.license.XPackLicenseState;
 import org.elasticsearch.tasks.Task;
@@ -106,7 +107,7 @@ private void explain(Task task, PutDataFrameAnalyticsAction.Request request,
         );
         if (licenseState.isSecurityEnabled()) {
             useSecondaryAuthIfAvailable(this.securityContext, () -> {
-                // Set the auth headers (preferring the secondary headers) to the caller's. 
+                // Set the auth headers (preferring the secondary headers) to the caller's.
                 // Regardless if the config was previously stored or not.
                 DataFrameAnalyticsConfig config = new DataFrameAnalyticsConfig.Builder(request.getConfig())
                     .setHeaders(filterSecurityHeaders(threadPool.getThreadContext().getHeaders()))
@@ -152,6 +153,11 @@ private void estimateMemoryUsage(Task task,
                                      DataFrameAnalyticsConfig config,
                                      ExtractedFields extractedFields,
                                      ActionListener<MemoryEstimation> listener) {
+        if (extractedFields.getAllFields().isEmpty()) {
+            listener.onResponse(new MemoryEstimation(ByteSizeValue.ZERO, ByteSizeValue.ZERO));
+            return;
+        }
+
         final String estimateMemoryTaskId = "memory_usage_estimation_" + task.getId();
         DataFrameDataExtractorFactory extractorFactory = DataFrameDataExtractorFactory.createForSourceIndices(
             new ParentTaskAssigningClient(client, task.getParentTaskId()), estimateMemoryTaskId, config, extractedFields);
diff --git a/x-pack/plugin/src/test/resources/rest-api-spec/test/ml/explain_data_frame_analytics.yml b/x-pack/plugin/src/test/resources/rest-api-spec/test/ml/explain_data_frame_analytics.yml
index 165ec26d4a43c..1d9ebf1909aa7 100644
--- a/x-pack/plugin/src/test/resources/rest-api-spec/test/ml/explain_data_frame_analytics.yml
+++ b/x-pack/plugin/src/test/resources/rest-api-spec/test/ml/explain_data_frame_analytics.yml
@@ -312,3 +312,35 @@
   - match: { field_selection.4.is_required: false }
   - match: { field_selection.4.feature_type: "categorical" }
   - is_false: field_selection.4.reason
+
+---
+"Test given no included field":
+
+  - do:
+      indices.create:
+        index: index-source
+        body:
+          mappings:
+            properties:
+              x:
+                type: keyword
+
+  - do:
+      index:
+        index: index-source
+        refresh: true
+        body: { x: "hello!" }
+  - match: { result: "created" }
+
+  - do:
+      ml.explain_data_frame_analytics:
+        body:
+          source: { index: "index-source" }
+          analysis: { outlier_detection: {} }
+  - match:
+      memory_estimation.expected_memory_without_disk: "0"
+  - match:
+      memory_estimation.expected_memory_with_disk: "0"
+  - length: { field_selection: 1 }
+  - match: { field_selection.0.name: "x" }
+  - match: { field_selection.0.is_included: false }