OpenSearch/docs/reference/release-notes/7.4.asciidoc

[[release-notes-7.4.2]]
== {es} version 7.4.2

[discrete]
[[bug-7.4.2]]
=== Bug fixes

Transform::
* Prevent assignment if any node is older than 7.4 {es-pull}48055[#48055] (issue: {es-issue}48019[#48019])

[[release-notes-7.4.1]]
== {es} version 7.4.1

Also see <<breaking-changes-7.4,Breaking changes in 7.4>>.

[[enhancement-7.4.1]]
[discrete]
=== Enhancements

Engine::
* Avoid unneeded refresh with concurrent realtime gets {es-pull}47895[#47895]
* sync before trimUnreferencedReaders to improve index preformance {es-pull}47790[#47790] (issues: {es-issue}46201[#46201], {es-issue}46203[#46203])
* Limit number of retaining translog files for peer recovery {es-pull}47414[#47414]

Infra/Circuit Breakers::
* Emit log message when parent circuit breaker trips {es-pull}47000[#47000]

Machine Learning::
* Throttle the delete-by-query of expired results {es-pull}47177[#47177] (issues: {es-issue}47003[#47003])
* The {ml} native processes are now arranged in a `.app` directory structure on
  macOS, to allow for notarization on macOS Catalina. {ml-pull}593[#593]



[[bug-7.4.1]]
[discrete]
=== Bug fixes

Aggregations::
* DocValueFormat implementation for date range fields {es-pull}47472[#47472] (issues: {es-issue}47323[#47323], {es-issue}47469[#47469])

Allocation::
* Dangling indices strip aliases {es-pull}47581[#47581]

Analysis::
* Reset Token position on reuse in `predicate_token_filter` {es-pull}47424[#47424] (issue: {es-issue}47197[#47197])

Authentication::
* Fix AD realm additional metadata {es-pull}47179[#47179] (issue: {es-issue}45848[#45848])

Authorization::
* Use 'should' clause instead of 'filter' when querying native privileges {es-pull}47019[#47019]

CCR::
* Do not auto-follow closed indices {es-pull}47721[#47721] (issue: {es-issue}47582[#47582])
* Relax maxSeqNoOfUpdates assertion in FollowingEngine {es-pull}47188[#47188] (issue: {es-issue}47137[#47137])

Cluster Coordination::
* Omit writing index metadata for non-replicated closed indices on data-only node {es-pull}47285[#47285] (issue: {es-issue}47276[#47276])

Features/ILM+SLM::
* Throw error retrieving non-existent SLM policy {es-pull}47679[#47679] (issue: {es-issue}47664[#47664])

Features/Indices APIs::
* Fix Rollover error when alias has closed indices {es-pull}47148[#47148] (issue: {es-issue}47146[#47146])

Features/Java High Level REST Client::
* Fix ILM HLRC Javadoc->Documentation links {es-pull}48083[#48083]

Features/Monitoring::
* [Monitoring] Add new cluster privilege now necessary for the stack monitoring ui {es-pull}47871[#47871]

Infra/Scripting::
* Drop stored scripts with the old style-id {es-pull}48078[#48078] (issue: {es-issue}47593[#47593])

Machine Learning::
* Fix detection of syslog-like timestamp in find_file_structure {es-pull}47970[#47970]
* Reinstate ML daily maintenance actions {es-pull}47103[#47103] (issue: {es-issue}47003[#47003])
* Fix possibility of crash when calculating forecasts that overflow to disk {ml-pull}688[#688]

Network::
* Fix es.http.cname_in_publish_address Deprecation Logging {es-pull}47451[#47451] (issue: {es-issue}47436[#47436])

SQL::
* SQL: Fix issue with negative literels and parentheses {es-pull}48113[#48113] (issue: {es-issue}48009[#48009])
* SQL: add "format" for "full" date range queries {es-pull}48073[#48073] (issue: {es-issue}48033[#48033])
* SQL: Allow whitespaces in escape patterns {es-pull}47577[#47577] (issue: {es-issue}47401[#47401])
* SQL: fix multi full-text functions usage with aggregate functions {es-pull}47444[#47444] (issue: {es-issue}47365[#47365])
* SQL: wrong number of values for columns {es-pull}42122[#42122]

Search::
* Fix alias field resolution in match query {es-pull}47369[#47369]

Snapshot/Restore::
* Fix Bug in Azure Repo Exception Handling {es-pull}47968[#47968]
* Fix Snapshot Corruption in Edge Case {es-pull}47552[#47552] (issues: {es-issue}46250[#46250], {es-issue}47550[#47550])

Store::
* Allow truncation of clean translog {es-pull}47866[#47866]

Transform::
* Fix bwc serialization with 7.3 {es-pull}48021[#48021]



[[release-notes-7.4.0]]
== {es} version 7.4.0

Also see <<breaking-changes-7.4,Breaking changes in 7.4>>.

[discrete]
=== Known issues

* Applying deletes or updates on an index after it has been shrunk may corrupt
the index. In order to prevent this issue, it is recommended to stop shrinking
read-write indices. For read-only indices, it is recommended to force-merge
indices after shrinking, which significantly reduces the likeliness of this
corruption in the case that deletes/updates would be applied by mistake. This
bug is fixed in {es} 7.7 and later versions. More details can be found on the
https://issues.apache.org/jira/browse/LUCENE-9300[corresponding issue].

* Activating the <<search-slow-log, search slow log>> should be avoided in this version.
Any attempt to log a slow search can throw an AIOOBE due to a bug that
performs concurrent modifications on a shared byte array.
(issue: {es-issue}/48358[#48358])

* Indices created in 6.x with <<date,`date`>> and <<date_nanos,`date_nanos`>> fields using formats
that are incompatible with java.time patterns will cause parsing errors, incorrect date calculations or wrong search results.
https://github.com/elastic/elasticsearch/pull/52555
This is fixed in {es} 7.7 and later versions.

[[breaking-7.4.0]]
[discrete]
=== Breaking changes

Infra/REST API::
* Update the schema for the REST API specification {es-pull}42346[#42346] (issue: {es-issue}35262[#35262])

Machine Learning::
* Improve progress reporting for data frame analytics {es-pull}45856[#45856]

Ranking::
* Forbid empty doc values on vector functions {es-pull}43944[#43944]

Search::
* Use float instead of double for query vectors. {es-pull}46004[#46004]

Snapshot/Restore::
* Provide an Option to Use Path-Style-Access with S3 Repo {es-pull}41966[#41966] (issue: {es-issue}41816[#41816])

Transforms::
* Combine task_state and indexer_state in _stats {es-pull}45276[#45276] (issue: {es-issue}45201[#45201])
* Improve response format of transform stats endpoint {es-pull}44350[#44350] (issue: {es-issue}43767[#43767])


[[breaking-java-7.4.0]]
[discrete]
=== Breaking Java changes

Geo::
* Geo: Change order of parameter in Geometries to lon, lat {es-pull}45332[#45332] (issue: {es-issue}45048[#45048])

Network::
* Stop Recreating Wrapped Handlers in RestController {es-pull}44964[#44964]



[[deprecation-7.4.0]]
[discrete]
=== Deprecations

Geo::
* Geo: add Geometry-based query builders to QueryBuilders {es-pull}45058[#45058] (issues: {es-issue}44715[#44715], {es-issue}45048[#45048])

Infra/Core::
* Bundle AdoptOpenJDK 13 {es-pull}46860[#46860]
* Add deprecation check for pidfile setting {es-pull}45939[#45939] (issues: {es-issue}45938[#45938], {es-issue}45940[#45940])
* Deprecate the pidfile setting {es-pull}45938[#45938]
* Add node.processors setting in favor of processors {es-pull}45855[#45855]
* Deprecate setting processors to more than available {es-pull}44889[#44889]

Infra/Settings::
* Add deprecation check for processors {es-pull}45925[#45925] (issues: {es-issue}45855[#45855], {es-issue}45905[#45905])

Machine Learning::
* Only emit deprecation warning if there was actual change of a datafeed's job_id. {es-pull}44755[#44755]
* Deprecate the ability to update datafeed's job_id. {es-pull}44691[#44691] (issue: {es-issue}44615[#44615])



[[feature-7.4.0]]
[discrete]
=== New features

Aggregations::
* Support Range Fields in Histogram and Date Histogram {es-pull}45395[#45395]
* Add Cumulative Cardinality agg (and Data Science plugin) {es-pull}43661[#43661] (issue: {es-issue}43550[#43550])

Analysis::
* Add support for inlined user dictionary in the Kuromoji plugin {es-pull}45489[#45489] (issue: {es-issue}25343[#25343])

Authentication::
* PKI realm authentication delegation {es-pull}45906[#45906] (issue: {es-issue}34396[#34396])
* PKI Authentication Delegation in new endpoint {es-pull}43796[#43796] (issue: {es-issue}34396[#34396])

Authorization::
* Add granular privileges for API keys {es-pull}42020[#42020]

Features/ILM::
* Add Snapshot Lifecycle Management {es-pull}43934[#43934] (issue: {es-issue}38461[#38461])

Features/Watcher::
* Add max_iterations configuration to watcher action with foreach execution {es-pull}45715[#45715] (issues: {es-issue}41997[#41997], {es-issue}45169[#45169])

Geo::
* [SPATIAL] New ShapeQueryBuilder for querying indexed cartesian geometry {es-pull}45108[#45108] (issue: {es-issue}44980[#44980])
* [GEO] New ShapeFieldMapper for indexing cartesian geometries {es-pull}44980[#44980]
* Add Circle Processor {es-pull}43851[#43851] (issue: {es-issue}43554[#43554])
* New `shape` field type for indexing Cartesian Geometries {es-pull}43644[#43644]

Machine Learning::
* Allow the user to specify 'query' in Evaluate Data Frame request {es-pull}45775[#45775] (issue: {es-issue}45729[#45729])
* Call the new _estimate_memory_usage API endpoint on data frame analytics _start {es-pull}45536[#45536] (issues: {es-issue}44699[#44699], {es-issue}45544[#45544])
* HLRC for memory usage estimation API {es-pull}45531[#45531] (issues: {es-issue}44699[#44699], {es-issue}45188[#45188])
* Implement ml/data_frame/analytics/_estimate_memory_usage API endpoint {es-pull}45188[#45188] (issue: {es-issue}44699[#44699])



[[enhancement-7.4.0]]
[discrete]
=== Enhancements

Aggregations::
* Add more flexibility to MovingFunction window alignment {es-pull}44360[#44360] (issue: {es-issue}42181[#42181])
* Optimize Min and Max BKD optimizations {es-pull}44315[#44315] (issue: {es-issue}44290[#44290])
* Allow pipeline aggs to select specific buckets from multi-bucket aggs {es-pull}44179[#44179]

Allocation::
* Defer reroute when starting shards {es-pull}44433[#44433] (issues: {es-issue}42105[#42105], {es-issue}42738[#42738])
* Allow RerouteService to reroute at lower priority {es-pull}44338[#44338]
* Auto-release of read-only-allow-delete block when disk utilization fa… {es-pull}42559[#42559] (issue: {es-issue}39334[#39334])

Analysis::
* Allow all token/char filters in normalizers {es-pull}43803[#43803] (issue: {es-issue}43758[#43758])

Authentication::
* Allow Transport Actions to indicate authN realm {es-pull}45767[#45767] (issue: {es-issue}45331[#45331])
* Explicitly fail if a realm only exists in keystore {es-pull}44471[#44471] (issue: {es-issue}44207[#44207])

Authorization::
* Add `manage_own_api_key` cluster privilege {es-pull}45897[#45897] (issue: {es-issue}40031[#40031])
* Consider `owner` flag when retrieving/invalidating keys with API key service {es-pull}45421[#45421] (issue: {es-issue}40031[#40031])
* REST API changes for manage-own-api-key privilege {es-pull}44936[#44936] (issue: {es-issue}40031[#40031])
* Simplify API key service API {es-pull}44935[#44935] (issue: {es-issue}40031[#40031])

CCR::
* Include leases in error message when operations no longer available {es-pull}45681[#45681]

CRUD::
* Return seq_no and primary_term for noop update {es-pull}44603[#44603] (issue: {es-issue}42497[#42497])

Cluster Coordination::
* Improve slow logging in MasterService {es-pull}45086[#45086] (issue: {es-issue}45007[#45007])
* More logging for slow cluster state application {es-pull}45007[#45007]
* Ignore unknown fields if overriding node metadata {es-pull}44689[#44689]
* Allow pending tasks before state recovery {es-pull}44685[#44685] (issue: {es-issue}44652[#44652])

Distributed::
* Do not create engine under IndexShard#mutex {es-pull}45263[#45263] (issue: {es-issue}43699[#43699])

Docs Infrastructure::
* add clarification around TESTSETUP docu and error message {es-pull}43306[#43306]

Engine::
* Flush engine after big merge {es-pull}46066[#46066]
* Do sync before closeIntoReader when rolling generation to improve index performance {es-pull}45765[#45765] (issue: {es-issue}45371[#45371])
* Refactor index engines to manage readers instead of searchers {es-pull}43860[#43860]
* Async IO Processor release before notify {es-pull}43682[#43682]
* Enable indexing optimization using sequence numbers on replicas {es-pull}43616[#43616] (issue: {es-issue}34099[#34099])

Features/ILM::
* Add node setting for disabling SLM {es-pull}46794[#46794] (issue: {es-issue}38461[#38461])
* Include in-progress snapshot for a policy with get SLM policy API {es-pull}45245[#45245]
*  Add option to filter ILM explain response {es-pull}44777[#44777] (issue: {es-issue}44189[#44189])
* Expose index age in ILM explain output {es-pull}44457[#44457] (issue: {es-issue}38988[#38988])

Features/Indices APIs::
* Add Clone Index API {es-pull}44267[#44267] (issue: {es-issue}44128[#44128])
* Add description to force-merge tasks {es-pull}41365[#41365] (issue: {es-issue}15975[#15975])

Features/Ingest::
* Fix IngestService to respect original document content type {es-pull}45799[#45799]
* Ingest Attachment: Upgrade tika to v1.22 {es-pull}45575[#45575]
* Retrieve processors instead of checking existence {es-pull}45354[#45354]
* Add ingest processor existence helper method {es-pull}45156[#45156]
* Change the ingest simulate api to not include dropped documents {es-pull}44161[#44161] (issue: {es-issue}36150[#36150])

Features/Java High Level REST Client::
* Add XContentType as parameter to HLRC ART#createServerTestInstance {es-pull}46036[#46036] (issue: {es-issue}45970[#45970])
* Add CloseIndexResponse to HLRC {es-pull}44349[#44349] (issue: {es-issue}39687[#39687])
* Add mapper-extras and the RankFeatureQuery in the hlrc {es-pull}43713[#43713] (issue: {es-issue}43634[#43634])
* removing background state update of Request object by RequestConverte… {es-pull}40156[#40156] (issue: {es-issue}39666[#39666])
* Add delete aliases API to the high-level REST client {es-pull}32909[#32909] (issue: {es-issue}27205[#27205])

Features/Watcher::
* Add SSL/TLS settings for watcher email {es-pull}45272[#45272] (issue: {es-issue}30307[#30307])
* Watcher reporting: add email warning if CSV attachment contains values that may be interperted as formulas {es-pull}44460[#44460]
* Watcher add stopped listener {es-pull}43939[#43939] (issue: {es-issue}42409[#42409])
* Improve CryptoService error message on missing secure file {es-pull}43623[#43623] (issue: {es-issue}43619[#43619])

Geo::
* Support WKT point conversion to geo_point type {es-pull}44107[#44107] (issue: {es-issue}41821[#41821])

Infra/Circuit Breakers::
* Fix G1 GC default IHOP {es-pull}46169[#46169]

Infra/Core::
* Add OCI annotations and adjust existing annotations {es-pull}45167[#45167] (issues: {es-issue}45162[#45162], {es-issue}45166[#45166])
* Use the full hash in build info {es-pull}45163[#45163] (issue: {es-issue}45162[#45162])

Infra/Packaging::
* Remove redundant Java check from Sys V init {es-pull}45793[#45793] (issue: {es-issue}45593[#45593])
* Notify systemd when Elasticsearch is ready {es-pull}44673[#44673]

Infra/Plugins::
* Make plugin verification FIPS 140 compliant {es-pull}44224[#44224] (issue: {es-issue}41263[#41263])

Infra/Scripting::
* Whitelist randomUUID in Painless {es-pull}45148[#45148] (issue: {es-issue}39080[#39080])
* Add missing ZonedDateTime methods for joda compat layer {es-pull}44829[#44829] (issue: {es-issue}44411[#44411])
* Remove stale permissions from untrusted policy {es-pull}44783[#44783]

Infra/Settings::
* Add more meaningful keystore version mismatch errors {es-pull}46291[#46291] (issue: {es-issue}44624[#44624])
* Lift the restrictions that uppercase is not allowed in Setting Name. {es-pull}45222[#45222] (issue: {es-issue}43835[#43835])
* Normalize environment paths {es-pull}45179[#45179] (issue: {es-issue}45176[#45176])

Machine Learning::
* Support boolean fields for data frame analytics {es-pull}46037[#46037]
* Add description to data frame analytics {es-pull}45774[#45774]
* Add regression analysis to data frame analytics {es-pull}45292[#45292]
* Introduce formal node ML role {es-pull}45174[#45174] (issues: {es-issue}29943[#29943], {es-issue}43175[#43175])
* Improve CSV header row detection in find_file_structure {es-pull}45099[#45099] (issue: {es-issue}45047[#45047])
* Outlier detection should only fetch docs that have the analyzed fields {es-pull}44944[#44944]
* Persist DatafeedTimingStats with RefreshPolicy.NONE by default {es-pull}44940[#44940] (issue: {es-issue}44792[#44792])
* Add result_type field to TimingStats and DatafeedTimingStats documents {es-pull}44812[#44812]
* Implement exponential average search time per hour statistics. {es-pull}44683[#44683] (issue: {es-issue}29857[#29857])
* Add r_squared eval metric to regression {es-pull}44248[#44248]
* Adds support for regression.mean_squared_error to eval API {es-pull}44140[#44140]
* Add DatafeedTimingStats.average_search_time_per_bucket_ms and TimingStats.total_bucket_processing_time_ms stats {es-pull}44125[#44125] (issue: {es-issue}29857[#29857])
* Add DatafeedTimingStats to datafeed GetDatafeedStatsAction.Response {es-pull}43045[#43045] (issue: {es-issue}29857[#29857])

Network::
* Better logging for TLS message on non-secure transport channel {es-pull}45835[#45835] (issue: {es-issue}32688[#32688])
* Asynchronously connect to remote clusters {es-pull}44825[#44825] (issue: {es-issue}40150[#40150])
* Improve errors when TLS files cannot be read {es-pull}44787[#44787] (issue: {es-issue}43079[#43079])
* Add per-socket keepalive options {es-pull}44055[#44055]
* Move ConnectionManager to async APIs {es-pull}42636[#42636]

Ranking::
* Search enhancement: pinned queries {es-pull}44345[#44345] (issue: {es-issue}44074[#44074])
* Fix parameter value for calling data.advanceExact {es-pull}44205[#44205]
* Add l1norm and l2norm distances for vectors {es-pull}44116[#44116] (issue: {es-issue}37947[#37947])

Recovery::
* Ignore translog retention policy if soft-deletes enabled {es-pull}45473[#45473] (issue: {es-issue}45136[#45136])
* Only retain reasonable history for peer recoveries {es-pull}45208[#45208] (issue: {es-issue}41536[#41536])
* Use index for peer recovery instead of translog {es-pull}45136[#45136] (issues: {es-issue}38904[#38904], {es-issue}41536[#41536], {es-issue}42211[#42211])
* Trim local translog in peer recovery {es-pull}44756[#44756]
* Make peer recovery send file chunks async {es-pull}44468[#44468] (issues: {es-issue}36195[#36195], {es-issue}44040[#44040])

SQL::
* SQL: Support queries with HAVING over SELECT {es-pull}46709[#46709] (issue: {es-issue}37051[#37051])
* SQL: Break TextFormatter/Cursor dependency {es-pull}45613[#45613] (issue: {es-issue}45516[#45516])
* SQL: remove deprecated use of "interval" from date_histogram usage {es-pull}45501[#45501] (issue: {es-issue}43922[#43922])
* SQL: use hasValue() methods from Elasticsearch's InspectionHelper classes {es-pull}44745[#44745] (issue: {es-issue}35745[#35745])
* Switch from using docvalue_fields to extracting values from _source {es-pull}44062[#44062] (issue: {es-issue}41852[#41852])

Search::
* Adds usage stats for vectors: {es-pull}44512[#44512]
* Associate sub-requests to their parent task in multi search API {es-pull}44492[#44492]
* Cancel search task on connection close {es-pull}43332[#43332]

Security::
* Set security index refresh interval to 1s {es-pull}45434[#45434] (issue: {es-issue}44934[#44934])

Snapshot/Restore::
* add disable_chunked_encoding configuration {es-pull}44052[#44052]
* Repository Cleanup Endpoint {es-pull}43900[#43900]

Task Management::
* Remove task null check in TransportAction {es-pull}45014[#45014]
* TaskListener#onFailure to accept Exception instead of Throwable {es-pull}44946[#44946]
* Move child task cancellation to TaskManager {es-pull}44573[#44573] (issue: {es-issue}44494[#44494])

Transforms::
* Add update transform api endpoint {es-pull}45154[#45154] (issue: {es-issue}43438[#43438])
* Add support for bucket_selector {es-pull}44718[#44718] (issues: {es-issue}43744[#43744], {es-issue}44557[#44557])
* Add force delete {es-pull}44590[#44590] (issue: {es-issue}43961[#43961])
* Add dynamic cluster setting for failure retries {es-pull}44577[#44577]
* Add optional defer_validation param to PUT {es-pull}44455[#44455] (issue: {es-issue}43439[#43439])
* Add support for geo_bounds aggregation {es-pull}44441[#44441]
* Add a frequency option to transform config, default 1m {es-pull}44120[#44120]


[[bug-7.4.0]]
[discrete]
=== Bug fixes

Aggregations::
* Fix early termination of aggregators that run with breadth-first mode {es-pull}44963[#44963] (issue: {es-issue}44909[#44909])
* Support BucketScript paths of type string and array. {es-pull}44694[#44694] (issue: {es-issue}44385[#44385])

Allocation::
* Avoid overshooting watermarks during relocation {es-pull}46079[#46079] (issue: {es-issue}45177[#45177])
* Cluster health should await events plus other things {es-pull}44348[#44348]
* Do not copy initial recovery filter during split {es-pull}44053[#44053] (issue: {es-issue}43955[#43955])

Analysis::
* Enable reloading of synonym_graph filters {es-pull}45135[#45135] (issue: {es-issue}45127[#45127])
* Fix AnalyzeAction response serialization {es-pull}44284[#44284] (issue: {es-issue}44078[#44078])

Authentication::
* Fallback to realm authc if ApiKey fails {es-pull}46538[#46538]
* Enforce realm name uniqueness {es-pull}46253[#46253]
* Allow empty token endpoint for implicit flow {es-pull}45038[#45038]
* Do not use scroll when finding duplicate API key {es-pull}45026[#45026]
* Fix broken short-circuit in getUnlicensedRealms {es-pull}44399[#44399]
* Fix X509AuthenticationToken principal {es-pull}43932[#43932] (issues: {es-issue}34396[#34396], {es-issue}43796[#43796])

Authorization::
* Do not rewrite aliases on remove-index from aliases requests {es-pull}46989[#46989]
* Give kibana user privileges to create APM agent config index {es-pull}46765[#46765] (issue: {es-issue}45610[#45610])
*  Add `manage_own_api_key` cluster privilege {es-pull}45696[#45696] (issue: {es-issue}40031[#40031])
* Sparse role queries can throw an NPE {es-pull}45053[#45053]

CCR::
* Clean up ShardFollowTasks for deleted indices {es-pull}44702[#44702] (issue: {es-issue}34404[#34404])
* Skip update if leader and follower settings identical {es-pull}44535[#44535] (issue: {es-issue}44521[#44521])
* Avoid stack overflow in auto-follow coordinator {es-pull}44421[#44421] (issue: {es-issue}43251[#43251])
* Avoid NPE when checking for CCR index privileges {es-pull}44397[#44397] (issue: {es-issue}44172[#44172])

CRUD::
* Ignore replication for noop updates {es-pull}46458[#46458] (issues: {es-issue}41065[#41065], {es-issue}44603[#44603], {es-issue}46366[#46366])
* Allow _update on write alias {es-pull}45318[#45318] (issue: {es-issue}31520[#31520])
* Do not allow version in Rest Update API {es-pull}43516[#43516] (issue: {es-issue}42497[#42497])

Cluster Coordination::
* Assert no exceptions during state application {es-pull}47090[#47090] (issue: {es-issue}47038[#47038])
* Avoid counting votes from master-ineligible nodes {es-pull}43688[#43688]

Distributed::
* Fix false positive out of sync warning in synced-flush {es-pull}46576[#46576] (issues: {es-issue}28464[#28464], {es-issue}30244[#30244])
* Suppress warning logs from background sync on relocated primary {es-pull}46247[#46247] (issues: {es-issue}40800[#40800], {es-issue}42241[#42241])
* Ensure AsyncTask#isScheduled remain false after close {es-pull}45687[#45687] (issue: {es-issue}45576[#45576])
* Update translog checkpoint after marking operations as persisted {es-pull}45634[#45634] (issue: {es-issue}29161[#29161])
* Fix clock used in update requests {es-pull}45262[#45262] (issue: {es-issue}45254[#45254])
* Restore DefaultShardOperationFailedException's reason during deserialization {es-pull}45203[#45203]
* Use IndicesModule named writables in elasticsearch-shard tool {es-pull}45036[#45036] (issue: {es-issue}44628[#44628])

Engine::
* Handle delete document level failures {es-pull}46100[#46100] (issue: {es-issue}46083[#46083])
* Handle no-op document level failures {es-pull}46083[#46083]
* Remove leniency during replay translog in peer recovery {es-pull}44989[#44989]
* Throw TranslogCorruptedException in more cases {es-pull}44217[#44217]
* Fail engine if hit document failure on replicas {es-pull}43523[#43523] (issues: {es-issue}40435[#40435], {es-issue}43228[#43228])

Features/ILM::
* Handle partial failure retrieving segments in SegmentCountStep {es-pull}46556[#46556]
* Fixes for API specification {es-pull}46522[#46522]
* Fix SnapshotLifecycleMetadata xcontent serialization {es-pull}46500[#46500] (issue: {es-issue}46499[#46499])
* Updated slm API spec parameters and URL {es-pull}44797[#44797]
* Fix swapped variables in error message {es-pull}44300[#44300]

Features/Indices APIs::
* Fix a bug where mappings are dropped from rollover requests. {es-pull}45411[#45411] (issue: {es-issue}45399[#45399])
* Create index with typeless mapping {es-pull}45120[#45120]
* Check shard limit after applying index templates {es-pull}44619[#44619] (issues: {es-issue}34021[#34021], {es-issue}44567[#44567], {es-issue}44619[#44619])
* Validate index settings after applying templates {es-pull}44612[#44612] (issues: {es-issue}34021[#34021], {es-issue}44567[#44567])

Features/Ingest::
* Allow dropping documents with auto-generated ID {es-pull}46773[#46773] (issue: {es-issue}46678[#46678])

Features/Java High Level REST Client::
* [HLRC] Send min_score as query string parameter to the count API {es-pull}46829[#46829] (issue: {es-issue}46474[#46474])
* HLRC multisearchTemplate forgot params {es-pull}46492[#46492] (issue: {es-issue}46488[#46488])
* terminateAfter added to the RequestConverter {es-pull}46474[#46474] (issue: {es-issue}46446[#46446])
* [Closes #44045] Added 'slices' parameter when submitting reindex request via Java high level REST client {es-pull}45690[#45690] (issue: {es-issue}44045[#44045])
* HLRC: Fix '+' Not Correctly Encoded in GET Req. {es-pull}33164[#33164] (issue: {es-issue}33077[#33077])

Features/Watcher::
* Fix class used to initialize logger in Watcher {es-pull}46467[#46467]
* Fix wrong URL encoding in watcher HTTP client {es-pull}45894[#45894] (issue: {es-issue}44970[#44970])
* Fix watcher HttpClient URL creation {es-pull}45207[#45207] (issue: {es-issue}44970[#44970])
* Log write failures for watcher history document. {es-pull}44129[#44129]

Geo::
* Geo: fix geo query decomposition {es-pull}44924[#44924] (issue: {es-issue}44891[#44891])
* Geo: add validator that only checks altitude {es-pull}43893[#43893]

Highlighting::
* Fix highlighting for script_score query {es-pull}46507[#46507] (issue: {es-issue}46471[#46471])

Infra/Core::
* Always check that cgroup data is present {es-pull}45606[#45606] (issue: {es-issue}45396[#45396])
* Safe publication of DelayedAllocationService and SnapshotShardsService {es-pull}45517[#45517] (issue: {es-issue}38560[#38560])
* Add default CLI JVM options {es-pull}44545[#44545] (issues: {es-issue}219[#219], {es-issue}42021[#42021])
* Fix decimal point parsing for date_optional_time {es-pull}43859[#43859] (issue: {es-issue}43730[#43730])

Infra/Logging::
* Fix types field in JSON Search Slow Logs {es-pull}44641[#44641]
* Add types field to JSON slow logs in 7.x {es-pull}44592[#44592] (issues: {es-issue}41354[#41354], {es-issue}44178[#44178])

Infra/Packaging::
* Add destructiveDistroTest meta task {es-pull}45762[#45762]
* Use bundled JDK in Sys V init {es-pull}45593[#45593] (issue: {es-issue}45542[#45542])
* Restore setting up temp dir for windows service {es-pull}44541[#44541]

Infra/Plugins::
* Do not checksum all bytes at once in plugin install {es-pull}44649[#44649] (issue: {es-issue}44545[#44545])

Infra/REST API::
* Improve error message when index settings are not a map {es-pull}45588[#45588] (issue: {es-issue}45126[#45126])
* Add is_write_index column to cat.aliases {es-pull}44772[#44772]
* Fix URL documentation in API specs {es-pull}44487[#44487]

Infra/Scripting::
* Fix bugs in Painless SCatch node {es-pull}45880[#45880]
* Fix JodaCompatibleZonedDateTime casts in Painless {es-pull}44874[#44874]

Infra/Settings::
* bug fix about elasticsearch.common.settings.Settings.processSetting {es-pull}44047[#44047] (issue: {es-issue}43791[#43791])

Machine Learning::
* Fix two datafeed flush lockup bugs {es-pull}46982[#46982]
* Throw an error when a datafeed needs CCS but it is not enabled for the node {es-pull}46044[#46044] (issue: {es-issue}46025[#46025])
* Handle "null" value of Estimate memory usage API response gracefully. {es-pull}45726[#45726] (issue: {es-issue}44699[#44699])
* Remove timeout on waiting for data frame analytics result processor to complete {es-pull}45724[#45724] (issue: {es-issue}45723[#45723])
* Check dest index is empty when starting data frame analytics {es-pull}45094[#45094]
* Catch any error thrown while closing data frame analytics process {es-pull}44958[#44958]
* Treat PostDataActionResponse.DataCounts.bucketCount as incremental rather than absolute (total). {es-pull}44803[#44803] (issue: {es-issue}44792[#44792])
* Treat big changes in searchCount as significant and persist the document after such changes {es-pull}44413[#44413] (issues: {es-issue}44196[#44196], {es-issue}44335[#44335])
* Update .ml-config mappings before indexing job, datafeed or data frame analytics config {es-pull}44216[#44216] (issue: {es-issue}44263[#44263])
* Wait for .ml-config primary before assigning persistent tasks {es-pull}44170[#44170] (issue: {es-issue}44156[#44156])
* Fix ML memory tracker lockup when inner step fails {es-pull}44158[#44158] (issue: {es-issue}44156[#44156])
* Fix datafeed checks when a concrete remote index is present {es-pull}43923[#43923] (issue: {es-issue}42113[#42113])
* Rename outlier detection method values `knn` and `tnn` to `distance_kth_nn` and `distance_knn`
respectively to match the API. {ml-pull}598[#598]
* Fix occasional (non-deterministic) reinitialisation of modeling for the `lat_long`
function. {ml-pull}641[#641]

Mapping::
* Make sure to validate the type before attempting to merge a new mapping. {es-pull}45157[#45157] (issues: {es-issue}29316[#29316], {es-issue}43012[#43012])
* Ensure field caps doesn't error on rank feature fields. {es-pull}44370[#44370] (issue: {es-issue}44330[#44330])
* Prevent types deprecation warning for indices.exists requests {es-pull}43963[#43963] (issue: {es-issue}43905[#43905])
* Add include_type_name in indices.exitst REST API spec {es-pull}43910[#43910] (issue: {es-issue}43905[#43905])

Network::
* Fix Broken HTTP Request Breaking Channel Closing {es-pull}45958[#45958] (issues: {es-issue}43362[#43362], {es-issue}43850[#43850])
* Fix plaintext on TLS port logging {es-pull}45852[#45852] (issue: {es-issue}32688[#32688])
* transport.publish_address should contain CNAME {es-pull}45626[#45626] (issues: {es-issue}32806[#32806], {es-issue}39970[#39970])
* Fix bug in copying bytes for socket write {es-pull}45463[#45463] (issue: {es-issue}45444[#45444])

Recovery::
* Never release store using CancellableThreads {es-pull}45409[#45409] (issues: {es-issue}45136[#45136], {es-issue}45237[#45237])
* Remove leniency in reset engine from translog {es-pull}44711[#44711]

Rollup::
* Fix Rollup job creation to work with templates {es-pull}43943[#43943]

SQL::
* SQL: Properly handle indices with no/empty mapping {es-pull}46775[#46775] (issue: {es-issue}46757[#46757])
* SQL: improve ResultSet behavior when no rows are available {es-pull}46753[#46753] (issue: {es-issue}46750[#46750])
* SQL: use the correct data type for types conversion {es-pull}46574[#46574] (issue: {es-issue}46090[#46090])
* SQL: Fix issue with common type resolution {es-pull}46565[#46565] (issue: {es-issue}46551[#46551])
* SQL: fix scripting for grouped by datetime functions {es-pull}46421[#46421] (issue: {es-issue}40241[#40241])
* SQL: Use null schema response {es-pull}46386[#46386] (issue: {es-issue}46381[#46381])
* SQL: Fix issue with IIF function when condition folds {es-pull}46290[#46290] (issue: {es-issue}46268[#46268])
* SQL: Fix issue with DataType for CASE with NULL {es-pull}46173[#46173] (issue: {es-issue}46032[#46032])
* SQL: adds format parameter to range queries for constant date comparisons {es-pull}45326[#45326] (issue: {es-issue}45139[#45139])
* SQL: uniquely named inner_hits sections for each nested field condition {es-pull}45039[#45039] (issues: {es-issue}33080[#33080], {es-issue}44544[#44544])
* SQL: fix URI path being lost in case of hosted ES scenario {es-pull}44776[#44776] (issue: {es-issue}44721[#44721])
* SQL: change the size of the list of concrete indices when resolving multiple indices {es-pull}43878[#43878] (issue: {es-issue}43876[#43876])
* SQL: handle double quotes escaping {es-pull}43829[#43829] (issue: {es-issue}43810[#43810])
* SQL: add pretty printing to JSON format {es-pull}43756[#43756]
* SQL: handle SQL not being available in a more graceful way {es-pull}43665[#43665] (issue: {es-issue}41279[#41279])

Search::
* Multi-get requests should wait for search active {es-pull}46283[#46283] (issue: {es-issue}27500[#27500])
* Ensure top docs optimization is fully disabled for queries with unbounded max scores. {es-pull}46105[#46105] (issue: {es-issue}45933[#45933])
* Disallow partial results when shard unavailable {es-pull}45739[#45739] (issue: {es-issue}42612[#42612])
* Prevent Leaking Search Tasks on Exceptions in FetchSearchPhase and DfsQueryPhase {es-pull}45500[#45500]
* Fix an NPE when requesting inner hits and _source is disabled. {es-pull}44836[#44836] (issue: {es-issue}43517[#43517])
* Don't use index_phrases on graph queries {es-pull}44340[#44340] (issue: {es-issue}43976[#43976])

Security::
* Initialize document subset bit set cache used for DLS {es-pull}46211[#46211] (issue: {es-issue}45147[#45147])
* Fix .security-* indices auto-create {es-pull}44918[#44918]
* SecurityIndexManager handle RuntimeException while reading mapping {es-pull}44409[#44409]
* Do not swallow I/O exception getting authentication {es-pull}44398[#44398] (issues: {es-issue}44172[#44172], {es-issue}44397[#44397])
* Use system context for looking up connected nodes {es-pull}43991[#43991] (issue: {es-issue}43974[#43974])

Snapshot/Restore::
* Fix Bug in Snapshot Status Response Timestamps {es-pull}46919[#46919] (issue: {es-issue}46913[#46913])
* GCS deleteBlobsIgnoringIfNotExists should catch StorageException {es-pull}46832[#46832] (issue: {es-issue}46772[#46772])
* Fix TransportSnapshotsStatusAction ThreadPool Use {es-pull}45824[#45824]
* Stop Executing SLM Policy Transport Action on Snapshot Pool {es-pull}45727[#45727] (issue: {es-issue}45594[#45594])
* Check again on-going snapshots/restores of indices before closing {es-pull}43873[#43873]
* Make Timestamps Returned by Snapshot APIs Consistent {es-pull}43148[#43148] (issue: {es-issue}43074[#43074])
* Recursively Delete Unreferenced Index Directories {es-pull}42189[#42189] (issue: {es-issue}13159[#13159])

Task Management::
* Catch AllocatedTask registration failures {es-pull}45300[#45300]

Transforms::
* Use field_caps API for mapping deduction {es-pull}46703[#46703] (issue: {es-issue}46694[#46694])
* Fix off-by-one error in checkpoint operations_behind {es-pull}46235[#46235]
* Moves failure state transition for MT safety {es-pull}45676[#45676] (issue: {es-issue}45664[#45664])
* Fix _start?force=true bug {es-pull}45660[#45660]
* Fix failure state transitions and race condition {es-pull}45627[#45627] (issues: {es-issue}45562[#45562], {es-issue}45609[#45609])
* Fix starting a batch data frame after stopping at runtime  {es-pull}45340[#45340] (issues: {es-issue}44219[#44219], {es-issue}45339[#45339])
* Fix null aggregation handling in indexer {es-pull}45061[#45061] (issue: {es-issue}44906[#44906])
* Unify validation exceptions between PUT and _preview {es-pull}44983[#44983] (issue: {es-issue}44953[#44953])
* Treat bulk index failures as an indexing failure {es-pull}44351[#44351] (issue: {es-issue}44101[#44101])
* Prevent task from attempting to run when failed {es-pull}44239[#44239] (issue: {es-issue}44121[#44121])
* Respond with 409 status code when failing _stop {es-pull}44231[#44231] (issue: {es-issue}44103[#44103])
* Add index validations to _start data frame transform {es-pull}44191[#44191] (issue: {es-issue}44104[#44104])
* Data frame task failure does not make a 500 response {es-pull}44058[#44058] (issue: {es-issue}44011[#44011])
* Audit message missing for autostop {es-pull}43984[#43984] (issue: {es-issue}43977[#43977])

[[regression-7.4.0]]
[discrete]
=== Regressions

Aggregations::
* Implement rounding optimization for fixed offset timezones {es-pull}46670[#46670] (issue: {es-issue}45702[#45702])



[[upgrade-7.4.0]]
[discrete]
=== Upgrades

Infra/Core::
* Update joda to 2.10.3 {es-pull}45495[#45495]

Infra/Packaging::
* Upgrade to JDK 12.0.2 {es-pull}45172[#45172]

Network::
* Upgrade to Netty 4.1.38 {es-pull}45132[#45132]

Search::
* Upgrade to lucene snapshot 8.3.0-snapshot-8dd116a6158 {es-pull}45604[#45604] (issue: {es-issue}43976[#43976])