Merge remote-tracking branch 'elastic/main' into faster-translog

Author: original-brownbear

docs/reference/search-connectors/release-notes.md

@@ -0,0 +1,36 @@
+---
+navigation_title: "Release notes"
+mapped_pages:
+  - https://www.elastic.co/guide/en/elasticsearch/reference/8.18/es-connectors-release-notes.html
+---
+
+# Connector release notes
+
+:::{admonition} Enterprise Search is discontinued in Elastic 9.0.0
+Please note that Enterprise Search is not available in Elastic 9.0+, including App Search, Workplace Search, the Elastic Web Crawler, and Elastic managed connectors.
+
+If you are an Enterprise Search user and want to upgrade to Elastic 9.0, refer to [our Enterprise Search FAQ](https://www.elastic.co/resources/search/enterprise-search-faq#what-features-are-impacted-by-this-announcement).
+It includes detailed steps, tooling, and resources to help you transition to supported alternatives in 9.x, such as Elasticsearch, the Open Web Crawler, and self-managed connectors.
+:::
+
+## 9.0.1 [connectors-9.0.1-release-notes]
+No changes since 9.0.0
+
+## 9.0.0 [connectors-9.0.0-release-notes]
+
+### Features and enhancements [connectors-9.0.0-features-enhancements]
+
+* Switched the default ingestion pipeline from `ent-search-generic-ingestion` to `search-default-ingestion`. The pipelines are functionally identical; only the name has changed to align with the deprecation of Enterprise Search. [#3049](https://github.com/elastic/connectors/pull/3049)
+* Removed opinionated index mappings and settings from Connectors. Going forward, indices will use Elastic’s default mappings and settings, rather than legacy App Search–optimized ones. To retain the previous behavior, create the index manually before pointing a connector to it. [#3013](https://github.com/elastic/connectors/pull/3013)
+
+### Fixes [connectors-9.0.0-fixes]
+
+* Fixed an issue where full syncs could delete newly ingested documents if the document ID from the third-party source was numeric. [#3031](https://github.com/elastic/connectors/pull/3031)
+* Fixed a bug where the Confluence connector failed to download some blog post documents due to unexpected response formats. [#2984](https://github.com/elastic/connectors/pull/2984)
+* Fixed a bug in the Outlook connector where deactivated users could cause syncs to fail. [#2967](https://github.com/elastic/connectors/pull/2967)
+* Resolved an issue where Network Drive connectors had trouble connecting to SMB 3.1.1 shares. [#2852](https://github.com/elastic/connectors/pull/2852)
+
+% ## Breaking changes [connectors-9.0.0-breaking-changes]
+% ## Deprications [connectorsch-9.0.0-deprecations]
+% ## Known issues [connectors-9.0.0-known-issues]
+

docs/reference/search-connectors/toc.yml

@@ -63,4 +63,5 @@ toc:
 - file: use-cases.md
   children:
   - file: es-connectors-overview-architecture.md
-- file: es-connectors-known-issues.md
+- file: es-connectors-known-issues.md
+- file: release-notes.md

muted-tests.yml

@@ -438,12 +438,6 @@ tests:
 - class: org.elasticsearch.xpack.remotecluster.RemoteClusterSecurityFcActionAuthorizationIT
   method: testIndicesPrivilegesAreEnforcedForCcrRestoreSessionActions
   issue: https://github.com/elastic/elasticsearch/issues/127782
-- class: org.elasticsearch.compute.aggregation.blockhash.BlockHashTests
-  method: test3BytesRefs {forcePackedHash=true}
-  issue: https://github.com/elastic/elasticsearch/issues/127825
-- class: org.elasticsearch.compute.aggregation.blockhash.BlockHashTests
-  method: test3BytesRefs {forcePackedHash=false}
-  issue: https://github.com/elastic/elasticsearch/issues/127826
 - class: org.elasticsearch.compute.aggregation.FilteredGroupingAggregatorFunctionTests
   method: testSimpleCircuitBreaking
   issue: https://github.com/elastic/elasticsearch/issues/127833

x-pack/plugin/esql/compute/src/test/java/org/elasticsearch/compute/aggregation/blockhash/BlockHashTests.java

@@ -46,7 +46,6 @@
 import java.util.stream.LongStream;
 
 import static org.hamcrest.Matchers.arrayWithSize;
-import static org.hamcrest.Matchers.either;
 import static org.hamcrest.Matchers.endsWith;
 import static org.hamcrest.Matchers.equalTo;
 import static org.hamcrest.Matchers.greaterThan;
@@ -1330,7 +1329,8 @@ public void close() {
                 Output o2 = output2.get(i);
                 assertThat(o1.offset, equalTo(o2.offset));
                 if (o1.vector != null) {
-                    assertThat(o1.vector, either(equalTo(o2.vector)).or(equalTo(o2.block.asVector())));
+                    assertNull(o1.block);
+                    assertThat(o1.vector, equalTo(o2.vector != null ? o2.vector : o2.block.asVector()));
                 } else {
                     assertNull(o2.vector);
                     assertThat(o1.block, equalTo(o2.block));

x-pack/plugin/esql/qa/testFixtures/src/main/resources/sample.csv-spec

@@ -4,7 +4,7 @@
 // range. These stats should be correctly adjusted for the sampling. Furthermore,
 // they also assert the value of MV_COUNT(VALUES(...)), which is not adjusted for
 // the sampling and therefore gives the size of the sample.
-// All ranges are very loose, so that the tests should fail less than 1 in a billion.
+// All ranges are very loose, so that the tests should practically never fail.
 // The range checks are done in ES|QL, resulting in one boolean value (is_expected),
 // because the CSV tests don't support such assertions.
 
@@ -40,10 +40,10 @@ required_capability: sample
 FROM employees
   | SAMPLE 0.5
   | STATS count = COUNT(), values_count = MV_COUNT(VALUES(emp_no)), avg_emp_no = AVG(emp_no), sum_emp_no = SUM(emp_no)
-  | EVAL is_expected = count >= 40 AND count <= 160 AND
-                       values_count >= 20 AND values_count <= 80 AND
+  | EVAL is_expected = count >= 20 AND count <= 180 AND
+                       values_count >= 10 AND values_count <= 90 AND
                        avg_emp_no > 10010 AND avg_emp_no < 10090 AND
-                       sum_emp_no > 40*10010 AND sum_emp_no < 160*10090
+                       sum_emp_no > 20*10010 AND sum_emp_no < 180*10090
   | KEEP is_expected
 ;
 
@@ -59,8 +59,8 @@ FROM employees
   | SAMPLE 0.5
   | WHERE emp_no > 10050 
   | STATS count = COUNT(), values_count = MV_COUNT(VALUES(emp_no)), avg_emp_no = AVG(emp_no)
-  | EVAL is_expected = count >= 10 AND count <= 90 AND 
-                       values_count >= 5 AND values_count <= 45 AND
+  | EVAL is_expected = count >= 5 AND count <= 95 AND
+                       values_count >= 2 AND values_count <= 48 AND
                        avg_emp_no > 10055 AND avg_emp_no < 10095
   | KEEP is_expected
 ;
@@ -77,8 +77,8 @@ FROM employees
   | WHERE emp_no <= 10050 
   | SAMPLE 0.5
   | STATS count = COUNT(), values_count = MV_COUNT(VALUES(emp_no)), avg_emp_no = AVG(emp_no)
-  | EVAL is_expected = count >= 10 AND count <= 90 AND 
-                       values_count >= 5 AND values_count <= 45 AND
+  | EVAL is_expected = count >= 5 AND count <= 95 AND
+                       values_count >= 2 AND values_count <= 48 AND
                        avg_emp_no > 10005 AND avg_emp_no < 10045
   | KEEP is_expected
 ;
@@ -95,8 +95,8 @@ FROM employees
   | SAMPLE 0.5
   | SORT emp_no
   | STATS count = COUNT(), values_count = MV_COUNT(VALUES(emp_no)), avg_emp_no = AVG(emp_no)
-  | EVAL is_expected = count >= 40 AND count <= 160 AND
-                       values_count >= 20 AND values_count <= 80 AND
+  | EVAL is_expected = count >= 20 AND count <= 180 AND
+                       values_count >= 10 AND values_count <= 90 AND
                        avg_emp_no > 10010 AND avg_emp_no < 10090
   | KEEP is_expected
 ;
@@ -113,8 +113,8 @@ FROM employees
   | SORT emp_no
   | SAMPLE 0.5
   | STATS count = COUNT(), values_count = MV_COUNT(VALUES(emp_no)), avg_emp_no = AVG(emp_no)
-  | EVAL is_expected = count >= 40 AND count <= 160 AND
-                       values_count >= 20 AND values_count <= 80 AND
+  | EVAL is_expected = count >= 20 AND count <= 180 AND
+                       values_count >= 10 AND values_count <= 90 AND
                        avg_emp_no > 10010 AND avg_emp_no < 10090
   | KEEP is_expected
 ;
@@ -147,8 +147,8 @@ FROM employees
   | LIMIT 50
   | SAMPLE 0.5
   | STATS count = COUNT(), values_count = MV_COUNT(VALUES(emp_no))
-  | EVAL is_expected = count >= 10 AND count <= 90 AND 
-                       values_count >= 5 AND values_count <= 45
+  | EVAL is_expected = count >= 5 AND count <= 95 AND
+                       values_count >= 2 AND values_count <= 48
   | KEEP is_expected
 ;
 
@@ -201,8 +201,8 @@ FROM employees
   | SAMPLE 0.8
   | SAMPLE 0.9
   | STATS count = COUNT(), values_count = MV_COUNT(VALUES(emp_no)), avg_emp_no = AVG(emp_no)
-  | EVAL is_expected = count >= 40 AND count <= 160 AND
-                       values_count >= 20 AND values_count <= 80 AND
+  | EVAL is_expected = count >= 20 AND count <= 180 AND
+                       values_count >= 10 AND values_count <= 90 AND
                        avg_emp_no > 10010 AND avg_emp_no < 10090
   | KEEP is_expected
 ;