[TestFix] ExplainLifecycleIT testStepInfoPreservedOnAutoRetry failing (#114294)

* Extend timeout of test and add logging on fail

* Unmute unstable test

* Switch to using logger for output

Keeps the forbiddenApis check happy

* Switch to using assertion messages to display

To display debug info

* Adjust logic of previous step info preservation

Add additional checks to ensure previous step info can't be cleared
when auto retrying, only updated with new info.

Also added logic to ensure previous step info is cleared when
transitioning to a new action

* Undo accidentally added lines from merge

Author: lukewhiting

x-pack/plugin/ilm/qa/multi-node/src/javaRestTest/java/org/elasticsearch/xpack/ilm/ExplainLifecycleIT.java

@@ -34,6 +34,7 @@
 import java.util.HashMap;
 import java.util.Locale;
 import java.util.Map;
+import java.util.concurrent.TimeUnit;
 
 import static org.elasticsearch.xcontent.XContentFactory.jsonBuilder;
 import static org.elasticsearch.xpack.TimeSeriesRestDriver.createFullPolicy;
@@ -307,14 +308,16 @@ public void testStepInfoPreservedOnAutoRetry() throws Exception {
 
         assertBusy(() -> {
             Map<String, Object> explainIndex = explainIndex(client(), indexName);
-            assertThat(explainIndex.get("failed_step_retry_count"), notNullValue());
-            assertThat(explainIndex.get("previous_step_info"), notNullValue());
-            assertThat((int) explainIndex.get("failed_step_retry_count"), greaterThan(0));
+            var assertionMessage = "Assertion failed for the following response: " + explainIndex;
+            assertThat(assertionMessage, explainIndex.get("failed_step_retry_count"), notNullValue());
+            assertThat(assertionMessage, explainIndex.get("previous_step_info"), notNullValue());
+            assertThat(assertionMessage, (int) explainIndex.get("failed_step_retry_count"), greaterThan(0));
             assertThat(
+                assertionMessage,
                 explainIndex.get("previous_step_info").toString(),
                 containsString("rollover_alias [" + aliasName + "] does not point to index [" + indexName + "]")
             );
-        });
+        }, 30, TimeUnit.SECONDS);
     }
 
     private void assertUnmanagedIndex(Map<String, Object> explainIndexMap) {

x-pack/plugin/ilm/src/main/java/org/elasticsearch/xpack/ilm/IndexLifecycleTransition.java

@@ -137,7 +137,8 @@ static ClusterState moveClusterStateToStep(
             lifecycleState,
             newStepKey,
             nowSupplier,
-            forcePhaseDefinitionRefresh
+            forcePhaseDefinitionRefresh,
+            true
         );
 
         return LifecycleExecutionStateUtils.newClusterStateWithLifecycleState(state, idxMeta.getIndex(), newLifecycleState);
@@ -175,6 +176,7 @@ static ClusterState moveClusterStateToErrorStep(
             currentState,
             new Step.StepKey(currentStep.phase(), currentStep.action(), ErrorStep.NAME),
             nowSupplier,
+            false,
             false
         );
 
@@ -243,7 +245,8 @@ static ClusterState moveClusterStateToPreviouslyFailedStep(
                 lifecycleState,
                 nextStepKey,
                 nowSupplier,
-                forcePhaseDefinitionRefresh
+                forcePhaseDefinitionRefresh,
+                false
             );
 
             LifecycleExecutionState.Builder retryStepState = LifecycleExecutionState.builder(nextStepState);
@@ -277,7 +280,8 @@ private static LifecycleExecutionState updateExecutionStateToStep(
         LifecycleExecutionState existingState,
         Step.StepKey newStep,
         LongSupplier nowSupplier,
-        boolean forcePhaseDefinitionRefresh
+        boolean forcePhaseDefinitionRefresh,
+        boolean allowNullPreviousStepInfo
     ) {
         Step.StepKey currentStep = Step.getCurrentStepKey(existingState);
         long nowAsMillis = nowSupplier.getAsLong();
@@ -289,7 +293,9 @@ private static LifecycleExecutionState updateExecutionStateToStep(
 
         // clear any step info or error-related settings from the current step
         updatedState.setFailedStep(null);
-        updatedState.setPreviousStepInfo(existingState.stepInfo());
+        if (allowNullPreviousStepInfo || existingState.stepInfo() != null) {
+            updatedState.setPreviousStepInfo(existingState.stepInfo());
+        }
         updatedState.setStepInfo(null);
         updatedState.setIsAutoRetryableError(null);
         updatedState.setFailedStepRetryCount(null);
@@ -390,7 +396,7 @@ public static LifecycleExecutionState moveStateToNextActionAndUpdateCachedPhase(
         updatedState.setStep(nextStep.name());
         updatedState.setStepTime(nowAsMillis);
         updatedState.setFailedStep(null);
-        updatedState.setPreviousStepInfo(existingState.stepInfo());
+        updatedState.setPreviousStepInfo(null);
         updatedState.setStepInfo(null);
         updatedState.setIsAutoRetryableError(null);
         updatedState.setFailedStepRetryCount(null);