Merge remote-tracking branch 'elastic/master' into remove-leftover-as…

…sertion

* elastic/master:
  Make packer cache branches explicit (elastic#41990)
  Re-mute all ml_datafeed_crud rolling upgrade tests
  Ensure testAckedIndexing uses disruption index settings
  add 7_3 as version (elastic#42368)
  Fix grammar problem in stemming reference. (elastic#42148)

.ci/packer_cache.sh

@@ -22,3 +22,4 @@ export JAVA8_HOME="${HOME}"/.java/java8
 export JAVA11_HOME="${HOME}"/.java/java11
 export JAVA12_HOME="${HOME}"/.java/openjdk12
 ./gradlew --parallel clean --scan -Porg.elasticsearch.acceptScanTOS=true -s resolveAllDependencies
+

distribution/bwc/build.gradle

@@ -239,12 +239,15 @@ bwcVersions.forPreviousUnreleased { BwcVersions.UnreleasedVersionInfo unreleased
 
         createBuildBwcTask(projectName, "${baseDir}/${projectName}", projectArtifact)
     }
-
 
     createRunBwcGradleTask("resolveAllBwcDependencies") {
         args 'resolveAllDependencies'
     }
-    resolveAllDependencies.dependsOn resolveAllBwcDependencies
+    Version currentVersion = Version.fromString(version)
+    if (currentVersion.getMinor() == 0 && currentVersion.getRevision() == 0) {
+        // We only want to resolve dependencies for live versions of master, without cascading this to older versions
+        resolveAllDependencies.dependsOn resolveAllBwcDependencies
+    }
 
     for (e in artifactFiles) {
         String projectName = e.key

docs/reference/how-to/recipes/stemming.asciidoc

@@ -171,7 +171,7 @@ the query need to be matched exactly while other parts should still take
 stemming into account?
 
 Fortunately, the `query_string` and `simple_query_string` queries have a feature
-that solve this exact problem: `quote_field_suffix`. This tell Elasticsearch
+that solves this exact problem: `quote_field_suffix`. This tells Elasticsearch
 that the words that appear in between quotes are to be redirected to a different
 field, see below:
 

server/src/main/java/org/elasticsearch/Version.java

@@ -92,6 +92,8 @@ public class Version implements Comparable<Version>, ToXContentFragment {
     public static final Version V_7_1_1 = new Version(V_7_1_1_ID, org.apache.lucene.util.Version.LUCENE_8_0_0);
     public static final int V_7_2_0_ID = 7020099;
     public static final Version V_7_2_0 = new Version(V_7_2_0_ID, org.apache.lucene.util.Version.LUCENE_8_0_0);
+    public static final int V_7_3_0_ID = 7030099;
+    public static final Version V_7_3_0 = new Version(V_7_3_0_ID, org.apache.lucene.util.Version.LUCENE_8_0_0);
     public static final int V_8_0_0_ID = 8000099;
     public static final Version V_8_0_0 = new Version(V_8_0_0_ID, org.apache.lucene.util.Version.LUCENE_8_1_0);
     public static final Version CURRENT = V_8_0_0;
@@ -110,6 +112,8 @@ public static Version fromId(int id) {
         switch (id) {
             case V_8_0_0_ID:
                 return V_8_0_0;
+            case V_7_3_0_ID:
+                return V_7_3_0;
             case V_7_2_0_ID:
                 return V_7_2_0;
             case V_7_1_1_ID:

server/src/main/java/org/elasticsearch/indices/recovery/PeerRecoveryTargetService.java

@@ -356,7 +356,7 @@ private StartRecoveryRequest getStartRecoveryRequest(final RecoveryTarget recove
             logger.trace("{} preparing for file-based recovery from [{}]", recoveryTarget.shardId(), recoveryTarget.sourceNode());
         } else {
             logger.trace(
-                "{} preparing for sequence-number-based recovery starting at local checkpoint [{}] from [{}]",
+                "{} preparing for sequence-number-based recovery starting at sequence number [{}] from [{}]",
                 recoveryTarget.shardId(),
                 startingSeqNo,
                 recoveryTarget.sourceNode());

server/src/test/java/org/elasticsearch/discovery/ClusterDisruptionIT.java

@@ -119,6 +119,7 @@ public void testAckedIndexing() throws Exception {
 
         assertAcked(prepareCreate("test")
             .setSettings(Settings.builder()
+                .put(indexSettings())
                 .put(IndexMetaData.SETTING_NUMBER_OF_SHARDS, 1 + randomInt(2))
                 .put(IndexMetaData.SETTING_NUMBER_OF_REPLICAS, randomInt(2))
             ));

x-pack/qa/rolling-upgrade/src/test/resources/rest-api-spec/test/mixed_cluster/40_ml_datafeed_crud.yml

@@ -1,3 +1,8 @@
+setup:
+  - skip:
+      version: "all"
+      reason:  "AwaitsFix https://github.com/elastic/elasticsearch/issues/42258"
+
 ---
 "Test old cluster datafeed without aggs":
   - do:

x-pack/qa/rolling-upgrade/src/test/resources/rest-api-spec/test/old_cluster/40_ml_datafeed_crud.yml

@@ -1,3 +1,8 @@
+setup:
+  - skip:
+      version: "all"
+      reason:  "AwaitsFix https://github.com/elastic/elasticsearch/issues/42258"
+
 ---
 "Put job and datafeed without aggs in old cluster":
 

x-pack/qa/rolling-upgrade/src/test/resources/rest-api-spec/test/upgraded_cluster/40_ml_datafeed_crud.yml

@@ -1,4 +1,8 @@
 setup:
+  - skip:
+      version: "all"
+      reason:  "AwaitsFix https://github.com/elastic/elasticsearch/issues/42258"
+
   - do:
      cluster.health:
         wait_for_status: green