You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Also see <<breaking-changes-8.15,Breaking changes in 8.15>>.
5
+
6
+
[[known-issues-8.15.1]]
7
+
[float]
8
+
=== Known issues
9
+
* Elasticsearch will not start if custom role mappings are configured using the
10
+
`xpack.security.authc.realms.*.files.role_mapping` configuration option. As a workaround, custom role mappings
11
+
can be configured using the https://www.elastic.co/guide/en/elasticsearch/reference/current/security-api-put-role-mapping.html[REST API] (issue: {es-issue}112503[#112503])
12
+
13
+
[[bug-8.15.1]]
14
+
[float]
15
+
=== Bug fixes
16
+
17
+
Aggregations::
18
+
* Revert "Avoid bucket copies in Aggs" {es-pull}111758[#111758] (issue: {es-issue}111679[#111679])
19
+
20
+
Authorization::
21
+
* Fix DLS over Runtime Fields {es-pull}112260[#112260] (issue: {es-issue}111637[#111637])
=== Stricter failure handling in multi-repo get-snapshots request handling
51
-
If a multi-repo get-snapshots request encounters a failure in one of the
52
-
targeted repositories then earlier versions of Elasticsearch would proceed
53
-
as if the faulty repository did not exist, except for a per-repository
54
-
failure report in a separate section of the response body. This makes it
55
-
impossible to paginate the results properly in the presence of failures. In
56
-
versions 8.15.0 and later this API's failure handling behaviour has been
57
-
made stricter, reporting an overall failure if any targeted repository's
58
-
contents cannot be listed.
59
-
60
-
{es-pull}107191[#107191]
61
-
62
-
[discrete]
63
-
[[add_new_int4_quantization_to_dense_vector]]
64
-
=== Add new int4 quantization to dense_vector
65
-
New int4 (half-byte) scalar quantization support via two knew index types: `int4_hnsw` and `int4_flat`.
66
-
This gives an 8x reduction from `float32` with some accuracy loss. In addition to less memory required, this
67
-
improves query and merge speed significantly when compared to raw vectors.
68
-
69
-
{es-pull}109317[#109317]
70
-
71
35
[discrete]
72
36
[[esql_inlinestats]]
73
37
=== ESQL: INLINESTATS
@@ -93,50 +57,6 @@ Produces output like:
93
57
94
58
{es-pull}109583[#109583]
95
59
96
-
[discrete]
97
-
[[mark_query_rules_as_ga]]
98
-
=== Mark Query Rules as GA
99
-
This PR marks query rules as Generally Available. All APIs are no longer
100
-
in tech preview.
101
-
102
-
{es-pull}110004[#110004]
103
-
104
-
[discrete]
105
-
[[adds_new_bit_element_type_for_dense_vectors]]
106
-
=== Adds new `bit` `element_type` for `dense_vectors`
107
-
This adds `bit` vector support by adding `element_type: bit` for
108
-
vectors. This new element type works for indexed and non-indexed
109
-
vectors. Additionally, it works with `hnsw` and `flat` index types. No
110
-
quantization based codec works with this element type, this is
111
-
consistent with `byte` vectors.
112
-
113
-
`bit` vectors accept up to `32768` dimensions in size and expect vectors
114
-
that are being indexed to be encoded either as a hexidecimal string or a
115
-
`byte[]` array where each element of the `byte` array represents `8`
116
-
bits of the vector.
117
-
118
-
`bit` vectors support script usage and regular query usage. When
119
-
indexed, all comparisons done are `xor` and `popcount` summations (aka,
120
-
hamming distance), and the scores are transformed and normalized given
121
-
the vector dimensions.
122
-
123
-
For scripts, `l1norm` is the same as `hamming` distance and `l2norm` is
124
-
`sqrt(l1norm)`. `dotProduct` and `cosineSimilarity` are not supported.
125
-
126
-
Note, the dimensions expected by this element_type are always to be
127
-
divisible by `8`, and the `byte[]` vectors provided for index must be
128
-
have size `dim/8` size, where each byte element represents `8` bits of
129
-
the vectors.
130
-
131
-
{es-pull}110059[#110059]
132
-
133
-
[discrete]
134
-
[[redact_processor_generally_available]]
135
-
=== The Redact processor is Generally Available
136
-
The Redact processor uses the Grok rules engine to obscure text in the input document matching the given Grok patterns. The Redact processor was initially released as Technical Preview in `8.7.0`, and is now released as Generally Available.
137
-
138
-
{es-pull}110395[#110395]
139
-
140
60
[discrete]
141
61
[[always_allow_rebalancing_by_default]]
142
62
=== Always allow rebalancing by default
@@ -149,76 +69,19 @@ version 8.16 `allow_rebalance` setting defaults to `always` unless the legacy al
149
69
150
70
{es-pull}111015[#111015]
151
71
152
-
// end::notable-highlights[]
153
-
154
-
155
-
[discrete]
156
-
[[new_custom_parser_for_iso_8601_datetimes]]
157
-
=== New custom parser for ISO-8601 datetimes
158
-
This introduces a new custom parser for ISO-8601 datetimes, for the `iso8601`, `strict_date_optional_time`, and
159
-
`strict_date_optional_time_nanos` built-in date formats. This provides a performance improvement over the
160
-
default Java date-time parsing. Whilst it maintains much of the same behaviour,
161
-
the new parser does not accept nonsensical date-time strings that have multiple fractional seconds fields
162
-
or multiple timezone specifiers. If the new parser fails to parse a string, it will then use the previous parser
163
-
to parse it. If a large proportion of the input data consists of these invalid strings, this may cause
164
-
a small performance degradation. If you wish to force the use of the old parsers regardless,
165
-
set the JVM property `es.datetime.java_time_parsers=true` on all ES nodes.
- Add a MemorySegment Vector scorer - for scoring without copying on-heap vectors: https://github.com/apache/lucene/pull/13339
202
-
203
-
{es-pull}109219[#109219]
83
+
{es-pull}111972[#111972]
204
84
205
-
[discrete]
206
-
[[synthetic_source_improvements]]
207
-
=== Synthetic `_source` improvements
208
-
There are multiple improvements to synthetic `_source` functionality:
209
-
210
-
* Synthetic `_source` is now supported for all field types including `nested` and `object`. `object` fields are supported with `enabled` set to `false`.
211
-
212
-
* Synthetic `_source` can be enabled together with `ignore_malformed` and `ignore_above` parameters for all field types that support them.
213
-
214
-
{es-pull}109501[#109501]
215
-
216
-
[discrete]
217
-
[[index_sorting_on_indexes_with_nested_fields]]
218
-
=== Index sorting on indexes with nested fields
219
-
Index sorting is now supported for indexes with mappings containing nested objects.
220
-
The index sort spec (as specified by `index.sort.field`) can't contain any nested
0 commit comments