| 6a37571 | ajrpayne | 03 December 2017, 14:20:10 UTC | Reflect changes in systemd service for LimitMEMLOCK The LimitMEMLOCK suggestion was removed from systemd service file and instead users should use an override file, so a comment in the environment file should be updated to reflect the same. Relates #27630 | 03 December 2017, 14:22:11 UTC |
| d42a848 | Luca Cavanna | 01 December 2017, 11:41:48 UTC | [DOCS] Clarify plan on transport client deprecation and removal (#27261) | 01 December 2017, 12:33:23 UTC |
| 95818e4 | Simon Willnauer | 30 November 2017, 19:44:05 UTC | Reset LiveVersionMap on sync commit (#27534) Today we carry on the size of the live version map to ensure that we minimze rehashing. Yet, once we are idle or we can issue a sync-commit we can resize it to defaults to free up memory. Relates to #27516 | 30 November 2017, 20:19:11 UTC |
| 68fed48 | Tanguy Leroux | 30 November 2017, 09:09:49 UTC | Do not swallow exception in ChecksumBlobStoreFormat.writeAtomic() (#27597) The ChecksumBlobStoreFormat.writeAtomic() method writes a blob using a temporary name and then moves the blob to its final name. The move operation can fail and in this case the temporary blob is deleted. If this delete operation also fails, then the initial exception is lost. This commit ensures that when something goes wrong during the move operation the initial exception is kept and thrown, and if the delete operation also fails then this additional exception is added as a suppressed exception to the initial one. | 30 November 2017, 09:25:29 UTC |
| 65af4c6 | Yannick Welsch | 30 November 2017, 09:03:16 UTC | Fix Gradle >=4.2 compatibility (#27591) Gradle 4.2 introduced a feature for safer handling of stale output files. Unfortunately, due to the way some of our tasks are written, this broke execution of our REST tests tasks. The reason for this is that the extract task (which extracts the ES distribution) would clean its output directory, and thereby also remove the empty cwd subdirectory which was created by the clean task. The reason why Gradle would remove the directory is that the earlier running clean task would programmatically create the empty cwd directory, but not make Gradle aware of this fact, which would result in Gradle believing that it could just safely clean the directory. This commit explicitly defines a task to create the cwd subdirectory, and marks the directory as output of the task, so that the subsequent extract task won't eagerly clean it. It thereby restores full compatibility of the build with Gradle 4.2 and above. This commit also removes the @Input annotation on the waitCondition closure of AntFixture, which conflicted with the extended input/output validation of Gradle 4.3. | 30 November 2017, 09:04:49 UTC |
| 38ecb4b | Jason Tedor | 30 November 2017, 00:57:26 UTC | Move DNS cache docs to system configuration docs When these docs were moved they should have been moved to the system configuration docs. This commit does that, and also fixes a missing heading that broke the docs build. | 30 November 2017, 01:00:10 UTC |
| 570696e | Jason Tedor | 29 November 2017, 23:02:26 UTC | Move DNS cache settings to important configuration This commit moves the DNS cache settings for the JVM to the important settings section of the docs. Relates #27592 | 29 November 2017, 23:04:29 UTC |
| f29219b | Jason Tedor | 28 November 2017, 16:29:47 UTC | Throw UOE from compressible bytes stream reset A compressible bytes output stream is a stream output which supports a reset method. However, compressible bytes output streams are unusual in that the current implementation sometimes supports a reset (if the stream is not compressed) and sometimes does not support a rest (if the stream is compressed). This inconsistent behavior is puzzling and instead we should simply always throw an unsupported operation exception. Relates #27564 | 29 November 2017, 14:13:22 UTC |
| e903468 | Jason Tedor | 27 November 2017, 18:48:04 UTC | Bubble exceptions when closing compressible streams Compressible bytes output stream swallows exceptions that occur when closing. This commit changes this behavior so that such exceptions bubble up. Relates #27542 | 29 November 2017, 14:13:09 UTC |
| de6ed75 | Jason Tedor | 29 November 2017, 14:09:20 UTC | Add CompressibleBytesOutputStream for compression This is a follow-up to #23941. Currently there are a number of complexities related to compression. The raw DeflaterOutputStream must be closed prior to sending bytes to ensure that EOS bytes are written. But the underlying ReleasableBytesStreamOutput cannot be closed until the bytes are sent to ensure that the bytes are not reused. Right now we have three different stream references hanging around in TCPTransport to handle this complexity. This commit introduces CompressibleBytesOutputStream to be one stream implemenation that will behave properly with or without compression enabled. Relates #27540 | 29 November 2017, 14:09:20 UTC |
| 7fd8494 | Tanguy Leroux | 29 November 2017, 08:39:49 UTC | Remove XContentType auto detection in BlobStoreRepository (#27480) | 29 November 2017, 08:49:05 UTC |
| 540edb9 | Jack Conradson | 28 November 2017, 22:49:59 UTC | Painless: Fix import statements. | 28 November 2017, 22:49:59 UTC |
| 025db5e | Jack Conradson | 28 November 2017, 21:44:52 UTC | Painless: Fix errors allowing void to be assigned to def. (#27460) | 28 November 2017, 22:47:38 UTC |
| 5e8e90f | Jim Ferenczi | 28 November 2017, 15:18:03 UTC | [TEST] AggregationsIntegrationIT#testScroll can timeout This change sets the scroll timeout for this test to 1m instead of 500ms in order to avoid loosing the scroll on slow machines. Relates #26378 | 28 November 2017, 15:21:06 UTC |
| 0ff3c04 | Jason Tedor | 25 November 2017, 00:00:18 UTC | Fix classes that can exit In a previous change, we locked down the classes that can exit by specifying explicit classes rather than packages than can exit. Alas, there was a bug in the sense that the class that we exit from in the case of an uncaught exception is not ElasticsearchUncaughtExceptionHandler but rather an anonymous nested class of ElasticsearchUncaughtExceptionHandler. To address this, we replace this anonymous class with a bonafide nested class ElasticsearchUncaughtExceptionHandler$PrivilegedHaltAction. Note that if we try to get this class name we have a $ in the middle of the string which is a special regular expression character; as such, we have to escape it. Relates #27518 | 25 November 2017, 00:01:25 UTC |
| c4c47d1 | Simon Willnauer | 24 November 2017, 13:57:31 UTC | Carry over version map size to prevent excessive resizing (#27516) Today we create a new concurrent hash map everytime we refresh the internal reader. Under defaults this isn't much of a deal but once the refresh interval is set to `-1` these maps grow quite large and it can have a significant impact on indexing throughput. Under low memory situations this can cause up to 2x slowdown. This change carries over the map size as the initial capacity wich will be auto-adjusted once indexing stops. Closes #20498 | 24 November 2017, 13:59:42 UTC |
| dc1b548 | Tanguy Leroux | 24 November 2017, 12:15:34 UTC | Delete shard store files before restoring a snapshot (#27476) Pull request #20220 added a change where the store files that have the same name but are different from the ones in the snapshot are deleted first before the snapshot is restored. This logic was based on the `Store.RecoveryDiff.different` set of files which works by computing a diff between an existing store and a snapshot. This works well when the files on the filesystem form valid shard store, ie there's a `segments` file and store files are not corrupted. Otherwise, the existing store's snapshot metadata cannot be read (using Store#snapshotStoreMetadata()) and an exception is thrown (CorruptIndexException, IndexFormatTooOldException etc) which is later caught as the begining of the restore process (see RestoreContext#restore()) and is translated into an empty store metadata (Store.MetadataSnapshot.EMPTY). This will make the deletion of different files introduced in #20220 useless as the set of files will always be empty even when store files exist on the filesystem. And if some files are present within the store directory, then restoring a snapshot with files with same names will fail with a FileAlreadyExistException. This is part of the #26865 issue. There are various cases were some files could exist in the store directory before a snapshot is restored. One that Igor identified is a restore attempt that failed on a node and only first files were restored, then the shard is allocated again to the same node and the restore starts again (but fails because of existing files). Another one is when some files of a closed index are corrupted / deleted and the index is restored. This commit adds a test that uses the infrastructure provided by IndexShardTestCase in order to test that restoring a shard succeed even when files with same names exist on filesystem. Related to #26865 | 24 November 2017, 13:14:56 UTC |
| 03990f4 | Martijn van Groningen | 22 November 2017, 10:07:29 UTC | fixed permission grant after securesm upgrade. | 22 November 2017, 10:08:26 UTC |
| 0408fd2 | Martijn van Groningen | 22 November 2017, 09:13:45 UTC | Bring back lost imports. | 22 November 2017, 09:13:45 UTC |
| 67e9151 | Jason Tedor | 21 November 2017, 21:46:10 UTC | Fix whitespace in Security.java | 21 November 2017, 21:49:23 UTC |
| e35b3c8 | Jason Tedor | 21 November 2017, 21:37:41 UTC | Tighten which classes can exit Today we allow exiting solely by being in certain packages. This commit upgrades the securesm dependency to a new version that supports being explicit about which classes can exit. We utilize that here to only allow exiting from the uncaught exception handler and the base CLI command class. Relates #27482 | 21 November 2017, 21:42:03 UTC |
| 000fb9b | Adrien Grand | 21 November 2017, 14:31:18 UTC | Fix dynamic mapping update generation. (#27467) When a field is not mapped, Elasticsearch tries to generate a mapping update from the parsed document. Some documents can introduce corner-cases, for instance in the event of a multi-valued field whose values would be mapped to different field types if they were supplied on their own, see for instance: ``` PUT index/doc/1 { "foo": ["2017-11-10T02:00:01.247Z","bar"] } ``` In that case, dynamic mappings want to map the first value as a `date` field and the second one as a `text` field. This currently throws an exception, which is expected, but the wrong one since it throws a `class_cast_exception` (which triggers a HTTP 5xx code) when it should throw an `illegal_argument_exception` (HTTP 4xx). | 21 November 2017, 14:53:32 UTC |
| e91182d | Ulrich Reffle | 21 November 2017, 10:10:35 UTC | [Docs] Fix broken bulleted lists (#27470) | 21 November 2017, 10:13:12 UTC |
| 7482686 | Dimitrios Liappis | 20 November 2017, 14:59:00 UTC | Tests: Add Fedora-27 to packaging tests Replace Fedora-25 with Fedora-27 and fix old references in `TESTING.asciidoc`. Relates #27434 | 20 November 2017, 15:06:36 UTC |
| ec02ec0 | K. Daniel Newton | 18 November 2017, 12:35:35 UTC | Correct usage of "an" to "a" in getting started docs This commit corrects a word usage error in the getting started docs. Since pronunciation is what determines when to use either "a" or "an" and the word "ubiquitous" is pronounced /yo͞oˈbikwədəs/, it should be preceded by "a." Relates #27420 | 18 November 2017, 12:36:54 UTC |
| ee15899 | Jason Tedor | 18 November 2017, 12:19:22 UTC | Avoid NPE when getting build information When the Elasticsearch code is loaded in an unusual classloading environment (e.g., when using the high-level REST client) in Jetty, the code source can be null and we trip with an NPE. This commit addresses this. Relates #27442 | 18 November 2017, 12:20:49 UTC |
| 4d117d4 | Simon Willnauer | 17 November 2017, 19:51:39 UTC | peanut butter hamburgers | 17 November 2017, 19:55:05 UTC |
| b4e462a | Yannick Welsch | 17 November 2017, 16:27:19 UTC | Fix place-holder in allocation decider messages (#27436) Allocation decider messages were using the wrong place-holder, which resulted in output of the form "no allocations are allowed due to {}" when showing diagnostics information in the explain API. | 17 November 2017, 16:28:42 UTC |
| 2d2715a | Jim Ferenczi | 16 November 2017, 11:02:09 UTC | [Docs] Add mapping.single_type in parent-join docs | 16 November 2017, 11:03:09 UTC |
| 3a2235f | Jim Ferenczi | 16 November 2017, 10:29:16 UTC | [Docs] Restore section about multi-level parent/child relation in parent-join (#27392) This section was removed to hide this ability to new users. This change restores the section and adds a warning regarding the expected performance. Closes #27336 | 16 November 2017, 11:03:09 UTC |
| 1f8550b | Martijn van Groningen | 26 October 2017, 09:25:10 UTC | docs: describe parent/child performances | 16 November 2017, 11:03:09 UTC |
| 285d137 | Jason Tedor | 15 November 2017, 18:33:59 UTC | Add note on plugin distributions in plugins folder This commit adds a note regarding not storing a plugin distribution in the plugins directory during installation or instllation will fail. Relates #27400 | 15 November 2017, 18:41:44 UTC |
| ceaa26b | Clinton Gormley | 15 November 2017, 15:50:34 UTC | Fixed remote.info test to use cluster.remote_info instead Closes #27396 | 15 November 2017, 15:50:34 UTC |
| 8f8216a | Tanguy Leroux | 15 November 2017, 14:30:27 UTC | Update Google SDK to version 1.23 (#27381) This commit updates the google-api-client library to version 1.23.0. Related to #26636 | 15 November 2017, 14:32:03 UTC |
| c76881e | Yannick Welsch | 15 November 2017, 13:50:13 UTC | Fix Gradle 4.3.1 compatibility for logging (#27382) The build currently does not work with Gradle 4.3.1 as the Gradle team stopped publishing the gradle-logging dependency to jcenter, starting with 4.3.1 (not sure why). There are two options: - Add the repository managed by Gradle team (https://repo.gradle.org/gradle/libs-releases-local) to our build - Use an older version (4.3) of the dependency when running with Gradle 4.3.1. Not to depend on another external repo, I've chosen solution 2. Note that this solution could break on future versions, but as this is a compileOnly dependency, and the interface we use has been super stable since forever, I don't envision this to be an issue (and easily detected by a breaking build). | 15 November 2017, 13:52:23 UTC |
| 1703c54 | Clinton Gormley | 14 November 2017, 12:33:23 UTC | REST: Renamed remote.info -> cluster.remote_info | 14 November 2017, 12:33:44 UTC |
| 0756216 | Clinton Gormley | 14 November 2017, 12:33:08 UTC | REST: Renamed ingest.processor.grok -> ingest.processor_grok | 14 November 2017, 12:33:44 UTC |
| ffd3906 | Tanguy Leroux | 14 November 2017, 10:43:32 UTC | Create new handlers for every new request in GoogleCloudStorageService (#27339) This commit changes the DefaultHttpRequestInitializer in order to make it create new HttpIOExceptionHandler and HttpUnsuccessfulResponseHandler for every new HTTP request instead of reusing the same two handlers for all requests. Closes #27092 | 14 November 2017, 11:08:41 UTC |
| bcc438a | Jason Tedor | 14 November 2017, 10:59:47 UTC | Revert "Reduce synchronization on field data cache" This reverts commit bda925764ae2f9b2bf55f6007f84f548f58c31fe. Relates #27365 | 14 November 2017, 10:59:47 UTC |
| 855d98b | Jason Tedor | 13 November 2017, 17:05:14 UTC | Upgrade AWS SDK Jackson Databind to 2.6.7.1 The AWS SDK has a transitive dependency on Jackson Databind. While the AWS SDK was recently upgraded, the Jackson Databind dependency was not pulled along with it to the version that the AWS SDK depends on. This commit upgrades the dependencies for discovery-ec2 and repository-s3 plugins to match versions on the AWS SDK transitive dependencies. Relates #27361 | 14 November 2017, 10:26:11 UTC |
| b963d42 | Tanguy Leroux | 13 November 2017, 13:22:14 UTC | Remove S3 output stream (#27280) Now the blob size information is available before writing anything, the repository implementation can know upfront what will be the more suitable API to upload the blob to S3. This commit removes the DefaultS3OutputStream and S3OutputStream classes and moves the implementation of the upload logic directly in the S3BlobContainer. related #26993 closes #26969 | 14 November 2017, 10:25:54 UTC |
| 5a5c581 | Tanguy Leroux | 10 November 2017, 16:05:32 UTC | Update to AWS SDK 1.11.223 (#27278) | 14 November 2017, 10:25:45 UTC |
| bda9257 | tinder-xli | 14 November 2017, 04:35:46 UTC | Reduce synchronization on field data cache The field data cache can come under heavy contention in cases when lots of search threads are hitting it for doc values. This commit reduces the amount of contention here by using a double-checked locking strategy to only lock when the cache needs to be initialized. Relates #27365 | 14 November 2017, 04:36:44 UTC |
| ea98113 | Yannick Welsch | 13 November 2017, 17:33:02 UTC | Fix compilation errors in IndexGraveyardTests backport gone wrong | 13 November 2017, 17:33:48 UTC |
| 08bcde1 | Yannick Welsch | 13 November 2017, 17:05:58 UTC | Properly format IndexGraveyard deletion date as date (#27362) The toXContent method for IndexGraveYard (which is a collection of tombstones for explicitly marking indices as deleted in the cluster state) confused timeValue with dateField, resulting in output of the form "delete_date" : "23424.3d" instead of "delete_date":"2017-11-13T15:50:51.614Z". | 13 November 2017, 17:07:04 UTC |
| 2bb6f83 | Yannick Welsch | 13 November 2017, 14:18:59 UTC | Stop responding to ping requests before master abdication (#27329) When the current master node is shutting down, it sends a leave request to the other nodes so that they can eagerly start a fresh master election. Unfortunately, it was still possible for the master node that was shutting down to respond to ping requests, possibly influencing the election decision as it still appeared as an active master in the ping responses. This commit ensures that UnicastZenPing does not respond to ping requests once it's been closed. ZenDiscovery.doStop() continues to ensure that the pinging component is first closed before it triggers a master election. Closes #27328 | 13 November 2017, 14:20:41 UTC |
| eb7b2b6 | Tanguy Leroux | 13 November 2017, 13:20:10 UTC | [Test] Fix POI version in packaging tests POI version has not been updated in packaging tests in #25003. Closes #27340 | 13 November 2017, 13:22:54 UTC |
| 3dd3c34 | tinder-xli | 11 November 2017, 03:28:58 UTC | Remove unnecessary logger creation for doc values field data This commit removes an unnecessary logger instance creation from the constructor for doc values field data. This construction is expensive for this oft-created class because of a synchronized block in the constructor for the logger. Relates #27349 | 11 November 2017, 03:31:13 UTC |
| d077078 | lcawley | 10 November 2017, 20:08:16 UTC | [DOCS] Fixed link to docker content | 10 November 2017, 20:09:08 UTC |
| ea664f5 | Lisa Cawley | 10 November 2017, 17:38:32 UTC | [DOCS] Move X-Pack-specific Docker content (#27333) | 10 November 2017, 18:54:55 UTC |
| 563a6c8 | Guillaume Le Floch | 09 November 2017, 18:16:44 UTC | Update Tika version to 1.15 This commit upgrades the Tika dependency to version 1.15. Relates #25003 | 09 November 2017, 18:18:39 UTC |
| 26c32d7 | Igor Motov | 03 November 2017, 23:36:08 UTC | Fix snapshot getting stuck in INIT state (#27214) If the master disconnects from the cluster after initiating snapshot, but just before the snapshot switches from INIT to STARTED state, the snapshot can get indefinitely stuck in the INIT state. This error is specific to v5.x+ and was triggered by keeping the master node that stepped down in the node list, the cleanup logic in snapshot/restore assumed that if master steps down it is always removed from the the node list. This commit changes the logic to trigger cleanup even if no nodes left the cluster. Closes #27180 | 08 November 2017, 20:52:22 UTC |
| 153b93f | Igor Motov | 08 November 2017, 15:43:28 UTC | Snapshot/Restore: better handle incorrect chunk_size settings in FS repo (#26844) Specifying a negative value or null as a chunk_size in FS repository can lead to corrupt snapshots. Closes #26843 | 08 November 2017, 20:52:22 UTC |
| 54085b0 | Jason Tedor | 08 November 2017, 20:22:31 UTC | Remove colons from task and configuration names Gradle 5.0 will remove support for colons in configuration and task names. This commit fixes this for our build by removing all current uses of colons in configuration and task names. Relates #27305 | 08 November 2017, 20:24:47 UTC |
| 94c9331 | Costin Leau | 08 November 2017, 14:42:45 UTC | Bumping ES Version to 5.6.5 | 08 November 2017, 14:42:45 UTC |
| 6667b36 | Boaz Leskes | 07 November 2017, 07:35:00 UTC | TemplateUpgradeService should only run on the master (#27294) The `TemplateUpgradeService` allows plugins to register a call back that mutates index templates upon recovery. This is handy for upgrade logic that needs to make sure that an existing index template is updated once the cluster is upgraded to a new version of the plugin (and ES). Currently, the service has complicated logic to decide which node should perform the upgrade. It will prefer the master node, if it is of the highest version of the cluster and otherwise it will fall back to one of the non-coordinating nodes which are on the latest version. While this attempts to make sure that new nodes can assume their template version is in place (but old node still need to be able to operate under both old and new template), it has an inherent problem in that the master (on an old version) may not be able to process the put template request with the new template - it may miss certain features. This PR changes the logic to be simpler and always rely on the current master nodes. This comes at the price that new nodes need to operate both with old templates and new. That price is small as they need to operate with old indices regardless of the template. On the flip side we reduce a lot of complexity in what can happen in the cluster. | 07 November 2017, 17:25:08 UTC |
| 3b76980 | Deb Adair | 07 November 2017, 15:45:08 UTC | [DOCS] Bumped version for 5.6.4 | 07 November 2017, 15:45:08 UTC |
| 77f234e | Jason Tedor | 07 November 2017, 12:34:14 UTC | Fix line length violations in engine test framework There are some line length violations in the engine test framework code EngineTestCase.java and TranslogHandler.java. This commit addresses them. | 07 November 2017, 12:35:02 UTC |
| d5508d7 | Clinton Gormley | 07 November 2017, 11:57:12 UTC | Added release notes for 5.6.4 | 07 November 2017, 11:57:12 UTC |
| 0251ff4 | Jason Tedor | 06 November 2017, 22:55:11 UTC | Die with dignity while merging If an out of memory error is thrown while merging, today we quietly rewrap it into a merge exception and the out of memory error is lost. Instead, we need to rethrow out of memory errors, and in fact any fatal error here, and let those go uncaught so that the node is torn down. This commit causes this to be the case. Relates #27265 | 06 November 2017, 23:59:03 UTC |
| fc09a09 | Jason Tedor | 06 November 2017, 18:59:31 UTC | Remove assumption from evil logging test This assumption could have been removed when Log4j was upgraded to version 2.9.1 yet it was left behind. This commit removes this unneeded assumption. | 06 November 2017, 19:03:22 UTC |
| fa0ad6d | Jason Tedor | 06 November 2017, 18:20:30 UTC | Correctly encode warning headers The warnings headers have a fairly limited set of valid characters (cf. quoted-text in RFC 7230). While we have assertions that we adhere to this set of valid characters ensuring that our warning messages do not violate the specificaion, we were neglecting the possibility that arbitrary user input would trickle into these warning headers. Thus, missing here was tests for these situations and encoding of characters that appear outside the set of valid characters. This commit addresses this by encoding any characters in a deprecation message that are not from the set of valid characters. Relates #27269 | 06 November 2017, 19:03:18 UTC |
| 7e64749 | Shubham Aggarwal | 06 November 2017, 18:15:36 UTC | Fixed references to Multi Index Syntax (#27283) | 06 November 2017, 18:19:02 UTC |
| 2b7ec5d | Patrice Bourgougnon | 06 November 2017, 17:13:27 UTC | Add an active Elasticsearch WordPress plugin link (#27279) | 06 November 2017, 17:15:24 UTC |
| 8159a66 | Russ Cam | 06 November 2017, 16:34:22 UTC | Align routing param type with search.json (#26958) Relates https://github.com/elastic/elasticsearch-net/issues/2869 | 06 November 2017, 16:40:13 UTC |
| 6c81489 | Boris Tyukin | 06 November 2017, 16:29:48 UTC | Update to support bulk updates by query (#27172) Getting started doc stated that bulk updates by query are not supported but they are now | 06 November 2017, 16:29:48 UTC |
| 3d46c58 | Pablo Musa | 06 November 2017, 09:16:52 UTC | [Docs] Fix minor paragraph indentation error for multiple Indices params (#25535) | 06 November 2017, 09:18:23 UTC |
| af20846 | olcbean | 06 November 2017, 09:11:25 UTC | Fix inconsistencies in the rest api specs for `tasks` (#27163) modify parameters names to reflect the changes done in the code base | 06 November 2017, 09:14:38 UTC |
| e0faf36 | Loek van Gool | 03 November 2017, 22:20:58 UTC | Add an example of dynamic field names (#27255) | 03 November 2017, 22:40:23 UTC |
| 95defa5 | David Pilato | 03 November 2017, 13:36:59 UTC | Add support for Gradle 4.3 (#27249) Closes #26840 Related to #27087 Backport of #27249 in 5.6 branch | 03 November 2017, 13:45:13 UTC |
| 4e07c3a | David Turner | 03 November 2017, 08:48:48 UTC | Reinstate recommendation for ≥ 3 master-eligible nodes. (#27204) In the docs for 1.7 ([doc][doc-1.7], [src][src-1.7]) there was a recommendation for at least 3 master-eligible nodes "in critical clusters" but this was lost when that page was updated in 2.0 ([doc][doc-2.0], [src][src-2.0]). I'd like to reinstate this. [doc-1.7]: https://www.elastic.co/guide/en/elasticsearch/reference/1.7/modules-node.html [src-1.7]: https://github.com/elastic/elasticsearch/blob/2cbaccb2f2a495923bc64447fe3396e0fc58b3d3/docs/reference/modules/node.asciidoc [doc-2.0]: https://www.elastic.co/guide/en/elasticsearch/reference/2.0/modules-node.html#split-brain [src-2.0]: https://github.com/elastic/elasticsearch/blob/4799009ad7ea8f885b6aedc6f62ad61d69e7a40d/docs/reference/modules/node.asciidoc | 03 November 2017, 10:57:41 UTC |
| 1fd5ac6 | Armin Braun | 03 November 2017, 00:22:19 UTC | Fixed byte buffer leak in Netty4 request handler If creating the REST request throws an exception (for example, because of invalid headers), we leak the request due to failure to release the buffer (which would otherwise happen after replying on the channel). This commit addresses this leak by handling the failure case. Relates #27222 | 03 November 2017, 00:23:31 UTC |
| 16e91a6 | Yannick Welsch | 02 November 2017, 15:52:10 UTC | Add additional explanations around discovery.zen.ping_timeout (#27231) Makes it clearer that this setting should only be changed with extra care. | 02 November 2017, 15:54:56 UTC |
| c103bd6 | Igor Motov | 01 November 2017, 00:04:00 UTC | Docs: restore now fails if it encounters incompatible settings (#26933) This change was introduced in 5.0.0, but the documentation wasn't updated to reflect it. Closes #26453 | 01 November 2017, 00:28:05 UTC |
| 7993de5 | zarucki | 31 October 2017, 22:24:01 UTC | Misleading incorrect for in loop. (#27195) | 31 October 2017, 22:24:01 UTC |
| 8bbedf5 | javanna | 31 October 2017, 14:24:46 UTC | [DOCS] Link remote info API in Cross Cluster Search docs page Closes #26327 | 31 October 2017, 14:26:52 UTC |
| a26062f | Jason Tedor | 26 October 2017, 02:05:46 UTC | Avoid stack overflow on search phases When a search is executing locally over many shards, we can stack overflow during query phase execution. This happens due to callbacks that occur after a phase completes for a shard and we move to the same phase on another shard. If all the shards for the query are local to the local node then we will never go async and these callbacks will end up as recursive calls. With sufficiently many shards, this will end up as a stack overflow. This commit addresses this by truncating the stack by forking to another thread on the executor for the phase. Relates #27069 | 31 October 2017, 13:56:40 UTC |
| 7fa51b3 | Jason Tedor | 13 September 2017, 10:16:27 UTC | Let search phases override max concurrent requests If the query coordinating node is also a data node that holds all the shards for a search request, we can end up recursing through the can match phase (because we send a local request and on response in the listener move to the next shard and do this again, without ever having returned from previous shards). This recursion can lead to stack overflow for even a reasonable number of indices (daily indices over a sixty days with five shards per day is enough to trigger the stack overflow). Moreover, all this execution would be happening on a network thread (the thread that initially received the query). With this commit, we allow search phases to override max concurrent requests. This allows the can match phase to avoid recursing through the shards towards a stack overflow. Relates #26484 | 31 October 2017, 13:56:40 UTC |
| 9188c9c | javanna | 31 October 2017, 13:41:44 UTC | Change 20_source_mapping test extension from yml to yaml and fix content The test was not run before, and because of this we didn't catch that one call was wrong. Fixed that as well. Closes #27126 | 31 October 2017, 13:41:44 UTC |
| 40fc8e4 | Martijn van Groningen | 31 October 2017, 09:09:44 UTC | Added deprecation logging about nested inner hits source response format change when source filtering is used. | 31 October 2017, 09:17:14 UTC |
| d407230 | Dimitrios Athanasiou | 30 October 2017, 15:19:43 UTC | [Docs] Fix note in bucket_selector | 30 October 2017, 15:22:09 UTC |
| 32226a5 | Clarkie | 30 October 2017, 10:53:57 UTC | [Docs] Fix indentation of examples (#27168) | 30 October 2017, 10:54:38 UTC |
| 7deed9e | Jim Ferenczi | 30 October 2017, 10:29:40 UTC | [Docs] Clarify `span_not` query behavior for non-overlapping matches (#27150) Closes #27134 | 30 October 2017, 10:31:53 UTC |
| bf61f52 | Christoph Büscher | 30 October 2017, 09:45:50 UTC | [Docs] Remove first person "I" from getting started (#27155) Avoid first person style and consistently switch to an unpersonal style in the getting started docs. | 30 October 2017, 09:48:01 UTC |
| e8c45e4 | Holger Bartnick | 30 October 2017, 08:31:55 UTC | [Docs] Correct link target for datatype murmur3 (#27143) | 30 October 2017, 08:34:34 UTC |
| dbc7f22 | Honza Král | 27 October 2017, 14:14:41 UTC | Exists template needs a template name (#25988) | 27 October 2017, 16:17:18 UTC |
| 07de031 | Christoph Büscher | 27 October 2017, 10:06:16 UTC | [Tests] Fix occasional test failure due to two random values being the same | 27 October 2017, 10:10:33 UTC |
| fa330a1 | Christoph Büscher | 27 October 2017, 08:07:36 UTC | Fix beidermorse phonetic token filter for unspecified `languageset` (#27112) Currently, when we create a BeiderMorseFilter with an unspecified `languageset`, the filter will not guess the language, which should be the default behaviour. This change fixes this and adds a simple test for the cases with and without provided `languageset` settings. Closes #26771 | 27 October 2017, 08:47:34 UTC |
| 9c56b41 | Jun Ohtani | 27 October 2017, 07:16:02 UTC | [Doc] Add Ingest CSV Processor Plugin to plugin as a community plugin (#27105) * [Doc] Add Ingest CSV Processor Plugin to plugin as a community plugin | 27 October 2017, 07:24:35 UTC |
| ffa0534 | Jun Ohtani | 27 October 2017, 02:52:04 UTC | [Doc] add restriction of analyze API (#27120) * [Doc] add restriction of no tokenize parameter Closes #26495 | 27 October 2017, 02:52:04 UTC |
| 60f3b49 | Jason Tedor | 26 October 2017, 16:33:37 UTC | Fix docs test for deleting a remote cluster This test in the docs shows how to delete a remote cluster. However, an unexpected deprecation header is returned. This commit addresses this by marking the deprecation header as expected. | 26 October 2017, 16:33:37 UTC |
| 6d34589 | İsmail Arılık | 26 October 2017, 09:05:47 UTC | [Docs] Fix wording for deleting child documents (#27066) | 26 October 2017, 09:05:47 UTC |
| 562d32a | Luca Cavanna | 25 October 2017, 11:26:06 UTC | Make ShardSearchTarget optional when parsing ShardSearchFailure (#27078) Turns out that `ShardSearchTarget` is nullable, hence its fields may not be printed out as part of `ShardSearchFailure#toXContent`, in which case `fromXContent` cannot parse it back. We would previously try to create the object with all of its fields set to null, but `Index` complains about it in the constructor. Also made sure that this code path is covered by our unit tests in `ShardSearchFailureTests`. Closes #27055 | 26 October 2017, 07:45:31 UTC |
| 8f3a538 | Jason Tedor | 26 October 2017, 02:36:53 UTC | Adjust .DS_Store test assertions on Windows Windows handles trying to read a file that does not exist because a component of the path is not a directory differently than other OS handle this situation. This commit adjusts these assertions for Windows. | 26 October 2017, 02:39:20 UTC |
| 8295649 | Jason Tedor | 26 October 2017, 02:15:38 UTC | Emit settings deprecation logging on empty update When executing a cluster settings update that leaves the cluster state unchanged, we skip validation and this avoids deprecation logging for deprecated settings in the cluster state. This commit addresses this by running validation even if the settings are unchanged. Relates #27017 | 26 October 2017, 02:16:40 UTC |
| 207158b | Jason Tedor | 25 October 2017, 15:25:29 UTC | Ignore .DS_Store files on macOS Finder creates these files if you browse a directory there. These files are really annoying, but it's an incredible pain for users that these files are created unbeknownst to them, and then they get in the way of Elasticsearch starting. This commit adds leniency on macOS only to skip these files. Relates #27108 | 25 October 2017, 15:27:55 UTC |
| 9f42848 | Loading Zhang | 25 October 2017, 14:15:10 UTC | Docs: Fix ingest geoip config location (#27110) | 25 October 2017, 14:15:10 UTC |
| 80f5cf6 | markwalkom | 25 October 2017, 10:38:50 UTC | [Docs] Clarify mapping `index` option default (#27104) | 25 October 2017, 10:38:50 UTC |
| 987d1a6 | Jason Tedor | 24 October 2017, 13:11:30 UTC | Blacklist Gradle 4.2 and above An upstream Gradle change has broken us starting on version 4.2. This commit blacklists these versions until we can either find a workaround, or the upstream issue is addressed. Relates #27087 | 24 October 2017, 13:12:45 UTC |
| 6eb5087 | Adrien Grand | 23 October 2017, 12:11:35 UTC | Reduce the default number of cached queries. (#26949) Memory usage of queries can't be properly accounted, which can be an issue when large queries are cached since the actual memory usage will be much higher than what the cache thinks. This problem is very hard if not impossible to fix so as a workaround I would like to decrease the maximum number of cached queries so that this problem is less likely to cause trouble in practice. For the record, this problem is more likely to occur in envirenments that have small shards or don't give much memory to the JVM. Closes #26938 | 23 October 2017, 12:24:35 UTC |
