Elasticsearch 0.90.8
See issues on GitHub
Release Notes
Breaking changes:
- Java client:
FilterBuilder
andQueryBuilder
throwElasticSearchIllegalArgumentException
on similar errors #4199
New features:
- Stats: Added
/_cluster/stats/
API #4460 - Fieldata: Unwanted loading of field data can now be prevented #4431
- Fieldata: Geo-Point field data can now be compressed #4386
- Query: Added support for Lucene's new
SimpleQueryParser
#4159 - Query: Term count on search results on a per-field basis #3920
- Highlighting: The
FastVectorHighlighter
now suppors combining of fields #3750
Enhancements:
- Upgrade to Lucene 4.6.0 #4241
- AllocationDeciders: Refactored
ShardRoutings
allocation/notification #4459 - AllocationDeciders: Improve RoutingNodes API visibility/mutability #4458
- AllocationDeciders: Deciders are executed in improved order, starting at cheap execution and most likely to return no #4454
- Field data: Configuration changes in field data settings are taken immediately into account #4430
- Field data: Added a field data based TermsFilter #4209
- Indices lifecycles:
IndicesLifecycle.Listener
to support listening for any index shard state change #4413 - Operation execution: Unify default ack timeout to 30 seconds #4395
- Mapping performance: Optimize dynamic mapping updates on master by processing latest one per index/node #4373
- Mapping: Allow field wildcards in the Get Field Mapping API #4367
- Mapping: Allow to disable sending a refresh-mapping to master node #4342
- Mapping: Move put mapping api to new acknowledgement mechanism #4228
- Mapping: Allow to add a date format to an existing mapping #3727
- Stats: Add IO operation stats to the File System statistics #4344
- Stats: Add a total section to file system stats #4343
- Stats: Added mlockall setting to process info output #4233
- Query:
SearchContext
pre allocates an exact sized array now #4156 - Query:
Multi_match
supports fields parameter as string #4164 - Query:
has_child
query can yield in wrong results if parent type has nested inner objects #4341 - Query: The
ignore_indices=missing
option also works for indices queries and filters #3428 - Query: Rename filter param to
post_filter
#4119 - Query: Deprecate
numeric_range
filter, add a field data execution option to range filter #4034 - Bulk: Bulk request should supports a timeout parameter similar to the index API #4220
- Scripting: Support primitive arrays as set values in a script #4175
- Open/Close Index API: Move to new acknowledgement mechanism #4169
- TransportClient:
TransportClient.connectedNodes
contains up to date node info#4162 - Aliases API: Move to new acknowledgement mechanism #4114
- Settings: Remove unused
term_index_interval/divisor
setting #3912 - Installation: Return error if elasticsearch was not built & packaged #2954
- Cluster management: On node join, evict existing node(s) with the same transport address #4503
- Testing: Add tests for REST layer #4469
- Stats: Add a new
usage
metric to CPU stats #4374 - Maven plugin updates
- Upgrade RPM Maven Plugin to 2.1-alpha-3 #4282
- Upgrade Maven Jar Plugin to 2.4 #4281
- Upgrade Maven Resources Plugin to 2.6 #4280
- Upgrade Maven Compiler Plugin to 3.1 #4279
- Upgrade Maven Assembly Plugin to 2.4 #4278
- Upgrade Maven Eclipse Plugin to 2.9 #4277
- Upgrade Maven Source Plugin to 2.2.1 #4276
- Upgrade Maven Surefire Plugin to 2.16 #4275
- Upgrade Maven Dependency Plugin to 2.8 #4274
- Update to shade plugin 2.2 to shade test artifact as well #4266
- OOM when building with java6 #4293
Bug fixes:
- Allocation: Don't delete local shard data when its allocation is tried on a non-existing node #4502
- Allocation: A cancelled allocation fails to reset the state of the source shard when moving #4457
- Allocation: Fixed accidental ignoring of calculated data in threshold check #4390
- Field data: Removed
IndexFieldData.getHighestNumberOfSeenUniqueValues()
as it was not accurate #4426 - Mapping: Some mappers are missing from
MapperBuilders
class #4423 - Mapping: Batch processing mapping updates can cause missed merged mappings when batching multiple types #4410
- Index templates: Loading templates via templates/ directory is not working #4235
- Index templates: Fixed loading index templates from
config/templates
directory using a specific syntax #4411 - Shutdown: Removed possible node deadlock #4334
- Boosting:
_all
field boosting was working only under specific conditions (as the default one) #4315 - Query: Query string caching could cause a matched filter to not return its name #4361
- Query: Apply slop correctly if phrase query is wrapped in a filtered query #4356
- Query:
has_parent
query could yield wrong results #4313 - Query:
has_child
filter and query did not take deleted documents correctly into account #4306 - Query:
min_score
is not inclusive #4303 - Query:
AllTermQuery
explanations have been corrected #4298 - Query:
has_child
query withscore_mode=avg
could emit infinity as score #4291 - Query: Fetch/Count operations might fail if executed on a relocated shard #4273
- Query: Indices filter doesn't support
_name
parameter #4166 - Query: Indices filter was applied to query irrelevant indices #2416
- Query: Return better error message when using key and value field in a range facet #4239
- Term suggest: Throw exception if different analysis chains (due to querying different indices for example) are tried to be used #3196
- Startup: During node startup local primaries should be preferred to relocating primaries #4237
- RobinEngine: Getting a search could result in a
NullPointerException
inRobinEngine.acquireSearcher()
#4232 - Translog: Fixed possible
NullPointerException
in FsTranslog when reverting transient log #4223 - Stats: Running the node stats api while a shard is moving onto the node logs an exception #4203
- Index warmers: Warmers API breaking when empty Warmup query is created #4196
- Plugins: Fixed an exception in
PluginsService
due to not having read permissions to the plugins directory #4186