Also see Breaking changes in 5.0.
Deprecationsedit
Enhancementsedit
-
Aggregations
-
-
Rescorer should be applied in the TopHits aggregation #20978 (issue: #19317)
-
Internal
-
-
ShardActiveResponseHandler shouldn’t hold to an entire cluster state #21470 (issue: #21394)
-
Network
-
-
Fix handler name on message not fully read #21478
-
Packaging
-
-
Add empty plugins dir for archive distributions #21204 (issue: #20342)
-
Make explicit missing settings for Windows service #21200 (issue: #18317)
-
Change permissions on config files #20966
-
Search
-
-
Optimize query with types filter in the URL (t/t/_search) #20979
Bug fixesedit
-
Aggregations
-
-
Rewrite Queries/Filter in FilterAggregationBuilder and ensure client usage marks query as non-cachable #21303 (issue: #21301)
-
Thread safety for scripted significance heuristics #21113 (issue: #18120)
-
ip_range aggregation should accept null bounds. #21043 (issue: #21006)
-
CAT API
-
-
Consume
full_id request parameter early #21270 (issue: #21266)
-
Cache
-
-
Fix the request cache keys to not hold references to the SearchContext. #21284
-
Circuit Breakers
-
-
ClusterState publishing shouldn’t trigger circuit breakers #20986 (issues: #20827, #20960)
-
Core
-
-
Fix ShardInfo#toString #21319
-
Protect BytesStreamOutput against overflows of the current number of written bytes. #21174 (issue: #21159)
-
Return target index name even if _rollover conditions are not met #21138
-
Engine
-
-
Retrying replication requests on replica doesn’t call
onRetry #21189 (issue: #20211)
-
Index APIs
-
-
Validate the
_rollover target index name early to also fail if dry_run=true #21330 (issue: #21149)
-
Ingest
-
-
Stored scripts and ingest node configurations should be included into a snapshot #21227 (issue: #21184)
-
Internal
-
-
Restore thread’s original context before returning to the ThreadPool #21411
-
Java API
-
-
Fix InternalSearchHit#hasSource to return the proper boolean value #21441 (issue: #21419)
-
Null checked for source when calling sourceRef #21431 (issue: #19279)
-
ClusterAdminClient.prepareDeletePipeline method should accept pipeline id to delete #21228
-
Java REST Client
-
-
Rest client: don’t reuse the same HttpAsyncResponseConsumer across multiple retries #21378
-
Logging
-
-
Assert status logger does not warn on Log4j usage #21339
-
Fix logger names for Netty #21223 (issue: #20457)
-
Packaging
-
-
Set vm.max_map_count on systemd package install #21507
-
Export ES_JVM_OPTIONS for SysV init #21445 (issue: #21255)
-
Debian: configure start-stop-daemon to not go into background #21343 (issues: #12716, #21300)
-
Generate POM files with non-wildcard excludes #21234 (issue: #21170)
-
Query DSL
-
-
Max score should be updated when a rescorer is used #20977 (issue: #20651)
-
REST
-
-
The routing query string param is supported by mget but was missing from the rest spec #21357
-
fix thread_pool_patterns path variable definition #21332
-
ensure the XContentBuilder is always closed in RestBuilderListener #21124
-
Reindex API
-
-
Bump reindex-from-remote’s buffer to 200mb #21222 (issue: #21185)
-
Fix reindex-from-remote for parent/child from <2.0 #21070 (issue: #21044)
-
Search
-
-
Fixes cachability problems with fetching TTL values when searching #21493 (issue: #21457)
-
Remove LateParsingQuery to prevent timestamp access after context is frozen #21328 (issue: #21295)
-
Snapshot/Restore
-
-
Ensures cleanup of temporary index-* generational blobs during snapshotting #21469 (issue: #21462)
-
Fixes get snapshot duplicates when asking for _all #21340 (issue: #21335)
-
Core
-
-
Upgrade to Lucene 6.2.1 #21207
-
Dates
-