257afc4 | Nik Everett | 18 April 2014, 21:01:29 UTC | Add support for conditional highlighting Adds a "conditional" object under each highlighting field that can contain other fields. Those fields are highlighted based on if there was a match in the containing field. It'll let you do things like this example in the docs: This example extracts extracts the first 100 characters from `text` if there is a match in `title`, otherwise it highlights `text` as normal. { "highlight": { "fields": { "title": { "conditional": { "match": { "text": { "no_match_size": 100, "skip_matching": true } } "no_match": { "text": { "no_match_size": 100 } }, } } } } } Also adds a setting to instruct the highlighter to skip its normal match logic and just do its no_match_size stuff. This is very useful with conditional highlighting "chains" that end in a no_match_size extracting a previous entry in the chain. Like this example from the docs: { "highlight": { "fields": { "text": { "conditional": { "no_match": { "auxiliary_text": { "conditional": { "no_match": { "file_text": { "conditional": { "no_match": { "text": { "no_match_size": 100, "skip_matching": true } } } } } } } } } } } } } Closes #4649 | 19 April 2014, 02:03:37 UTC |
12bbe28 | Boaz Leskes | 17 April 2014, 09:12:36 UTC | Fail replica shards locally upon failures When a replication operation (index/delete/update) fails to be executed properly, we fail the replica and allow master to allocate a new copy of it. At the moment, the node hosting the primary shard is responsible of notifying the master of a failed replica. However, if the replica shard is initializing (`POST_RECOVERY` state), we have a racing condition between the failed shard message and moving the shard into the `STARTED` state. If the latter happen first, master will fail to resolve the fail shard message. This commit builds on #5800 and fails the engine of the replica shard if a replication operation fails. This protects us against the above as the shard will reject the `STARTED` command from master. It also makes us more resilient to other racing conditions in this area. Closes #5847 | 18 April 2014, 16:56:08 UTC |
b6515e2 | Simon Willnauer | 18 April 2014, 16:20:26 UTC | [TEST] Make InternalEngineMergeTests more stable | 18 April 2014, 16:20:44 UTC |
ff6b234 | Sean Gallagher | 17 April 2014, 20:17:51 UTC | Removed leading spaces on commented config key lines in elasticsearch.yml Author: Sean Gallagher Date: 17 Apr 2014 16:18 EDT Removed spaces on commented lines containing config key entries to prevent users from inadvertently messing up the indents in elasticsearch.yml. Closes #5842 | 18 April 2014, 15:54:52 UTC |
442dda2 | javanna | 18 April 2014, 13:07:15 UTC | [TEST] _id is not indexed by default, sort on score,_uid in MultiMatchQueryTests | 18 April 2014, 13:09:00 UTC |
a808fe9 | Martijn van Groningen | 17 April 2014, 10:18:43 UTC | Moved the updateMappingOnMaster logic into a single place. Closes #5798 | 18 April 2014, 12:27:13 UTC |
d6a6767 | javanna | 18 April 2014, 10:15:27 UTC | [TEST] added sort by "_id" when score is the same to MultiMatchQueryTests#testEquivalence A merge (and refresh) might rarely happen in the background between the two queries whose output is compared. It might then happen that two docs with same scores get returned by the two queries in a different order due to different lucene document id (which has changed in the meantime). To fix this we need to order by id when the score is the same, so that we can safely compare the output of the two queries (multimatch and dismax). | 18 April 2014, 10:15:44 UTC |
4b5e8ce | Xiao Yu | 15 April 2014, 20:15:14 UTC | Add a site plugin into list Howdy, Not sure if this is kosher but I would like to add my site plugin to the list in the docs. | 17 April 2014, 17:28:37 UTC |
eaf04fc | Simon Willnauer | 16 April 2014, 19:51:53 UTC | [BUILD] Fail release if test have AwaitsFix annotations Note this commit upgrades Forbidden APIs to version 1.5 Closes #5807 | 17 April 2014, 14:45:46 UTC |
e3e631e | Christoph Frick | 02 April 2014, 21:01:12 UTC | Update allocation.asciidoc | 17 April 2014, 12:42:58 UTC |
a73286b | Martijn van Groningen | 17 April 2014, 04:51:11 UTC | [TEST] Use startNodesAsync in unicast discovery tests. | 17 April 2014, 04:51:11 UTC |
4c30277 | Igor Motov | 16 April 2014, 21:28:43 UTC | [DOCS] Make snapshot repository examples consistent | 16 April 2014, 21:28:43 UTC |
0948260 | Simon Willnauer | 16 April 2014, 20:59:31 UTC | [TEST] make testTimeoutSendExceptionWithDelayedResponse more reliable on slow systems | 16 April 2014, 20:59:31 UTC |
1755ae7 | Simon Willnauer | 16 April 2014, 15:21:08 UTC | Added version constants for 1.1.2 and 1.0.4 | 16 April 2014, 15:21:19 UTC |
0887e68 | Boaz Leskes | 16 April 2014, 13:59:56 UTC | [Test] InternalEngineTests: increased gc deletes interval & turn it off randomly | 16 April 2014, 13:59:56 UTC |
65906d1 | Clinton Gormley | 16 April 2014, 13:41:06 UTC | Update multi-match-query.asciidoc Typo | 16 April 2014, 13:41:38 UTC |
26adb37 | Simon Willnauer | 16 April 2014, 13:17:38 UTC | [TEST] Ignore bogus system properties. LuceneTestCase might reset some solr properties that cause our tests to fail if the run before in the same JVM We just ignore solr properties. | 16 April 2014, 13:19:17 UTC |
3530c8b | Simon Willnauer | 16 April 2014, 13:08:52 UTC | [TEST] catch exceptions if TTL already expired when indexing TTLPercolatorTests indexes docs with small TTLs which can trigger AlreadyExpiredException exception. This is expected while rare and we should just catch them. | 16 April 2014, 13:10:28 UTC |
be14968 | Simon Willnauer | 14 April 2014, 14:37:41 UTC | Ensure close is called under lock in the case of an engine failure Until today we did close the engine without aqcuireing the write lock since most calls were still holding a read lock. This commit removes the code that holds on to the readlock when failing the engine which means we can simply call #close() | 16 April 2014, 12:50:40 UTC |
099b9c6 | Boaz Leskes | 16 April 2014, 12:45:54 UTC | add debug logs if failed shards can not be resolved. | 16 April 2014, 12:45:54 UTC |
840d1b4 | Martijn van Groningen | 16 April 2014, 08:49:01 UTC | [TEST] Reduce the amount of docs being indexed. | 16 April 2014, 08:49:24 UTC |
98deb55 | Martijn van Groningen | 09 April 2014, 05:48:35 UTC | Better deal with invalid scroll ids. Closes #5738 | 16 April 2014, 07:13:29 UTC |
8df5d4c | Simon Willnauer | 15 April 2014, 20:50:33 UTC | [TEST] Fix PercolatorTests#testSimple2 This test requires a mapping since otherwise if there is no mapping added the percolator query might not be parsed as a query on a numeric field since the query might arrive on a node before the dynamic mapping reached that node. This commit also moves the `indexService.readAllowed()` call up before the number of percolation queries is check to make sure we fail if reads are not allowed - there might be a query in-flight which means we need to check another node rather than return an empty result. | 15 April 2014, 21:01:35 UTC |
65e72a5 | Lee Hinman | 15 April 2014, 17:18:51 UTC | [TEST] Wait for green, and refresh after indexing in percolator test | 15 April 2014, 17:19:41 UTC |
de59cde | Kouhei Sutou | 15 April 2014, 04:55:39 UTC | Remove garbage | 15 April 2014, 15:57:25 UTC |
c5c87c4 | Simon Willnauer | 15 April 2014, 13:06:33 UTC | [TEST] Don't delete data dirs after test - only delete their content. Closes #5815 | 15 April 2014, 15:03:31 UTC |
9898eed | Simon Willnauer | 15 April 2014, 13:23:25 UTC | [DOCS] Update merge docs to reflect the max_merge_at_once property | 15 April 2014, 14:42:23 UTC |
320a206 | Simon Willnauer | 15 April 2014, 13:14:44 UTC | Switch back to ConcurrentMergeScheduler Load tests showed that SerialMS has problems to keep up with the merges under high load. We should switch back to CMS until we have a better story to balance merge threads / efforts across shards on a single node. Closes #5817 | 15 April 2014, 14:42:23 UTC |
9920084 | Adrien Grand | 15 April 2014, 14:20:16 UTC | [TEST] Wait for shards to be allocated before running testUpdateMappingDynamicallyWhilePercolating. If the percolate request is executed soon enough, all shards fail and the mapping is not actually updated. | 15 April 2014, 14:20:16 UTC |
9ea0e3a | Scott Wilkerson | 30 March 2014, 19:03:54 UTC | Update percolate.asciidoc fix typo | 15 April 2014, 14:01:44 UTC |
c61110c | eliasah | 15 April 2014, 13:56:38 UTC | Update core-types.asciidoc Missing bracket | 15 April 2014, 13:57:04 UTC |
d7fda62 | Yousef | 13 April 2014, 12:28:24 UTC | Updated date_formats to new dynamic_date_formats | 15 April 2014, 13:44:08 UTC |
202b1e2 | Martijn van Groningen | 11 April 2014, 12:14:22 UTC | Update clusterstate if mapping service has local changes If the during percolating a new field was introduced in the local mapping service, then those changes should be updated in cluster state of the master as well. Closes #5776 | 15 April 2014, 11:41:01 UTC |
7c6d745 | Simon Willnauer | 14 April 2014, 19:45:28 UTC | Cleanup FileSystemUtils#mkdirs(File) This methods had some workarounds for bugs that seem to be fixed in Java 7 [1]. There seem to be other problems on shared file-systems which are not really supported by lucene anyway or rather not recommeded. Yet the current solution that interrupts a static thread reference is too dangrous given all the usage of NIO across elasticsearch. [1] http://bugs.java.com/bugdatabase/view_bug.do?bug_id=4742723 | 15 April 2014, 11:22:51 UTC |
8dd5dd4 | Simon Willnauer | 14 April 2014, 19:39:13 UTC | Remove FileSystemUtils#maxOpenFiles This method basically forcefully creates as many files as possible to find out the process limit in a brute-force manner. The number of possible probles with this approach would exceed the number of lines left on this commit message. This commit uses a JMX based alternative to print the process limit. | 15 April 2014, 11:22:51 UTC |
bc5bdbc | Shay Banon | 15 April 2014, 11:17:20 UTC | Remove jsr166y now that we on Java 7, cleanup jsr166e to classes we use | 15 April 2014, 11:17:28 UTC |
8bede70 | Simon Willnauer | 10 April 2014, 15:27:05 UTC | Use TransportBulkAction for internal request from IndicesTTLService This prevents executing bulks internal autocreate indices logic and ensures that this internal request never creates an index automaticall. This fixes a bug where the TTL purger thread ran after the actual index it was purging was already closed / deleted and that re-created that index. Closes #5766 | 15 April 2014, 10:40:25 UTC |
7ef36d9 | Andrew Selden | 15 April 2014, 00:36:43 UTC | Separate benchmark API endpoints Separates benchmark API endpoints into separate files according to API funtionality. This makes it easier for our tests and clients. Closes #5787 | 15 April 2014, 00:36:43 UTC |
3d23a71 | Igor Motov | 13 April 2014, 22:45:37 UTC | Fix snapshot status with empty repository The snapshot status command with empty repository should return current status of currently running snapshots in all repositories. Fixes #5790 | 14 April 2014, 23:02:41 UTC |
2cf66c4 | Andrew Selden | 14 April 2014, 21:08:41 UTC | Benchmark documentation Moving benchmark documentation under the search section. Closes #5786 | 14 April 2014, 21:08:41 UTC |
2ed8c63 | Igor Motov | 31 March 2014, 18:26:45 UTC | Separate persistent and global metadata serialization settings | 14 April 2014, 20:25:33 UTC |
27d4d76 | Simon Willnauer | 14 April 2014, 20:22:51 UTC | [BUILD] remove leftover print statement | 14 April 2014, 20:23:10 UTC |
0564c88 | Simon Willnauer | 09 April 2014, 14:15:15 UTC | Remove unused FileSystemUtils#copyFile | 14 April 2014, 19:48:27 UTC |
a215dd3 | Simon Willnauer | 09 April 2014, 14:01:00 UTC | Prevent fsync from creating 0-byte files This is related to LUCENE-5570 where fsync creates a 0-byte file if the file does not exists. This commit adds the patched lucene version using Java 7 APIs as well as a note to replace this method with the upcomeing IOUtils#fsync in Lucene 4.8 This commit cleans up FsImmutableBlobContainer#writeBlob to make use of Java7 Auto-Closing features and ensures that the directory the blob was written to is fsynced as well if possible. | 14 April 2014, 19:48:23 UTC |
11bf13c | Simon Willnauer | 14 April 2014, 16:24:42 UTC | Check for no open issues before build release | 14 April 2014, 16:54:04 UTC |
754eb16 | Simon Willnauer | 09 April 2014, 09:24:33 UTC | Upgrade to Lucene 4.7.2 | 14 April 2014, 16:30:03 UTC |
e458d4f | Adrien Grand | 11 April 2014, 09:18:49 UTC | Improved SearchContext.addReleasable. For resources that have their life time effectively defined by the search context they are attached to, it is convenient to use the search context to schedule the release of such resources. This commit changes aggregations to use this mechanism and also introduces a `Lifetime` object that can be used to define how long the object should live: - COLLECTION: if the object only needs to live during collection time and is what SearchContext.addReleasable would have chosen before this change (used for p/c queries), - SEARCH_PHASE for resources that only need to live during the current search phase (DFS, QUERY or FETCH), - SEARCH_CONTEXT for resources that need to live until the context is destroyed. Aggregators are currently registed with SEARCH_CONTEXT. The reason is that when using the DFS_QUERY_THEN_FETCH search type, they are allocated during the DFS phase but only used during the QUERY phase. However we should fix it in order to only allocate them during the QUERY phase and use SEARCH_PHASE as a life time. Close #5703 | 14 April 2014, 15:42:41 UTC |
e589301 | Adrien Grand | 04 April 2014, 09:50:39 UTC | Make Releasable extend AutoCloseable. Java7's AutoCloseable allows to manage resources more nicely using try-with-resources statements. Since the semantics of our Releasable interface are very close to a Closeable, let's switch to it. Close #5689 | 14 April 2014, 15:21:42 UTC |
e688f44 | Adrien Grand | 14 April 2014, 10:31:34 UTC | [TEST] Use indexRandom in ShardSizeTests. | 14 April 2014, 10:31:34 UTC |
1ce56ff | Simon Willnauer | 14 April 2014, 10:15:02 UTC | Revert "Don't lookup version for auto generated id and create" This reverts commit dc734984541537dc93ddae4633ad69898bdb6959. | 14 April 2014, 10:15:02 UTC |
dc73498 | Shay Banon | 12 April 2014, 10:10:56 UTC | Don't lookup version for auto generated id and create When a create document is executed, and its an auto generated id (based on UUID), we know that the document will not exists in the index, so there is no need to try and lookup the version from the index. For many cases, like logging, where ids are auto generated, this can improve the indexing performance, specifically for lightweight documents where analysis is not a big part of the execution. | 14 April 2014, 08:06:53 UTC |
f853718 | Peter Dyson | 04 April 2014, 01:29:51 UTC | [DOCS] update old status of plugins | 14 April 2014, 00:18:19 UTC |
ad143e1 | Simon Willnauer | 12 April 2014, 21:02:14 UTC | [TEST] Fix ClusterStatsTests#testValuesSmokeScreen to wait for yellow to get reliable FS stats. | 12 April 2014, 21:02:31 UTC |
ec3c635 | Simon Willnauer | 12 April 2014, 19:54:46 UTC | [TEST] use a real upperbound for the check on the time spend during suggestions | 12 April 2014, 19:54:46 UTC |
e9c0dd9 | Shay Banon | 12 April 2014, 14:14:58 UTC | [Test] should be abstract | 12 April 2014, 14:14:58 UTC |
efb7499 | Simon Willnauer | 11 April 2014, 20:46:44 UTC | [TEST] Improve performance of MockBigArray MockPageRecycler | 11 April 2014, 21:02:59 UTC |
5d611a9 | Simon Willnauer | 11 April 2014, 13:40:07 UTC | Ensure pending merges are updated on segment flushes Due to the default of `async_merge` to `true` we never run the merge policy on a segment flush which prevented the pending merges from being updated and that caused actual pending merges not to contribute to the merge decision. This commit also removes the `index.async.merge` setting is actually misleading since we take care of merges not being excecuted on the indexing threads on a different level (the merge scheduler) since 1.1. This commit also adds an additional check when to run a refresh since soely relying on the dirty flag might leave merges un-refreshed which can cause search slowdowns and higher memory consumption. Closes #5779 | 11 April 2014, 21:02:59 UTC |
e0fbd5d | Boaz Leskes | 11 April 2014, 17:50:40 UTC | PR #5706 introduced a bug in the sparse array-backed field data When we load sparse single valued data, we automatically assign a missing value to represent a document who has none. We try to find a value that will increase the number of bits needed to represent the data. If that missing value happen to be 0, we do no properly intialize the value array. This commit solved this problem but also cleans up the code even more to make spotting such issues easier in the future. | 11 April 2014, 19:34:36 UTC |
63d1fa4 | Boaz Leskes | 11 April 2014, 16:12:35 UTC | Added awaitFix for SimpleNestedTests.testSortNestedWithNestedFilter While investigating failures | 11 April 2014, 16:12:35 UTC |
8ce3bba | Malte Schirnacher | 10 April 2014, 13:35:05 UTC | Fix typos in percolate.asciidoc Close #5762 #5763 #5764 | 11 April 2014, 16:09:16 UTC |
f549472 | Boaz Leskes | 11 April 2014, 14:40:47 UTC | Fixed- PackedArrayIndexFieldData.chooseStorageFormat compared to Long.MAX_VALUE instead of Long.MIN_VALUE Also made the LongFieldDataTests.SINGLE_VALUED_SPARSE_RANDOM & LongFieldDataTests.MULTI_VALUED_SPARSE_RANDOM more sparse | 11 April 2014, 14:40:47 UTC |
1d1ca3b | Boaz Leskes | 27 March 2014, 19:43:36 UTC | Added a AppendingDeltaPackedLongBuffer-based storage format to single value field data The AppendingDeltaPackedLongBuffer uses delta compression in paged fashion. For data which is roughly monotonic this results in reduced memory signature. By default we use the storage format expected to use the least memory. You can force a choice using a new field data setting `memory_storage_hint` which can be set to `ORDINALS`, `PACKED` or `PAGED` Closes #5706 | 11 April 2014, 13:50:34 UTC |
e8ea9d7 | Chris Earle | 31 March 2014, 04:02:32 UTC | Strengthening pseudo random number generator and adding tests to verify its behavior. Closes #5454 and #5578 | 11 April 2014, 12:01:40 UTC |
45a1b44 | Martijn van Groningen | 11 April 2014, 10:33:35 UTC | Each search request should use a new InternalSearchResponse instance even in case when all shards return no hits. The InternalSearchResponse may get modified afterwards, so a new instance required at all times. | 11 April 2014, 10:44:21 UTC |
862611b | Simon Willnauer | 11 April 2014, 07:03:28 UTC | [TEST] Prevent TTLPurger from recreating deleted index Related to #5766 | 11 April 2014, 07:03:28 UTC |
01794bf | Martijn van Groningen | 11 April 2014, 06:47:41 UTC | Ignored clear scroll rest test | 11 April 2014, 06:47:41 UTC |
b400129 | Martijn van Groningen | 08 April 2014, 13:41:32 UTC | The clear scroll apis now optionally accepts a scroll_id in it body. Closes #5726 | 11 April 2014, 03:39:59 UTC |
c4a49c2 | Martijn van Groningen | 08 April 2014, 08:56:12 UTC | Fixed an where a nested `nested` aggregation falls outside of its parent `nested` aggregation bounds. The nested `nested` aggs now gets the proper parent docs that define its bounds correctly. Closes #5728 | 11 April 2014, 03:16:17 UTC |
80ebd49 | Sean Gallagher | 10 April 2014, 19:23:07 UTC | [DOCS] Added tables and fixes to upgrade.asciidoc, fixed version in README.textile Author: Sean Gallagher Date: 10 Apr 2014 15:23 EDT | 10 April 2014, 19:23:07 UTC |
40f1913 | Nik Everett | 10 April 2014, 17:32:34 UTC | [Docs] Add experimental highlighter plugin | 10 April 2014, 17:32:34 UTC |
7bf3ffe | Martijn van Groningen | 01 April 2014, 10:07:22 UTC | Added test with AwaitsFix annotation that simulates a split brain. | 10 April 2014, 13:13:59 UTC |
7a26b49 | Shay Banon | 10 April 2014, 12:59:11 UTC | Use startNodesAsync in more tests | 10 April 2014, 12:59:11 UTC |
53cbaa0 | Adrien Grand | 10 April 2014, 09:34:41 UTC | [TESTS] Add more assertSearchResponse to aggregations tests. | 10 April 2014, 09:34:41 UTC |
2067996 | Simon Willnauer | 10 April 2014, 08:49:32 UTC | [TEST] Allow higher ClusterHealthTimeout on tests if configuration is slow | 10 April 2014, 08:49:32 UTC |
bceac03 | Simon Willnauer | 09 April 2014, 21:16:52 UTC | Fix _bench endpoint specs to have the body on the right level in the json object | 09 April 2014, 21:16:52 UTC |
34a3e8a | Simon Willnauer | 09 April 2014, 21:16:17 UTC | Add better error reporting to RestApiParser if assertions are tripped | 09 April 2014, 21:16:17 UTC |
e2c8ff9 | Andrew Selden | 12 March 2014, 16:16:39 UTC | Benchmark API Add an API endpoint at /_bench for submitting, listing, and aborting search benchmarks. This API can be used for timing search requests, subject to various user-defined settings. Benchmark results provide summary and detailed statistics on such values as min, max, and mean time. Values are reported per-node so that it is easy to spot outliers. Slow requests are also reported. Long running benchmarks can be viewed with a GET request, or aborted with a POST request. Benchmark results are optionally stored in an index for subsequent analysis. Closes #5407 | 09 April 2014, 20:06:55 UTC |
af0278b | Nik Everett | 09 April 2014, 17:53:24 UTC | [Docs] Allocation setting explanation Closes #5748 | 09 April 2014, 18:11:36 UTC |
9aa1cb4 | Lee Hinman | 08 April 2014, 16:44:50 UTC | Fix format string for DiskThresholdDecider reroute explanation | 09 April 2014, 14:49:49 UTC |
960d353 | Costin Leau | 17 March 2014, 13:17:49 UTC | Remove plugin isolation feature for a future version relates #5261 | 09 April 2014, 14:28:11 UTC |
a9c8624 | Simon Willnauer | 09 April 2014, 09:45:40 UTC | [TEST] MLT Rest test needs a mapping since we randomized number of nodes/shards | 09 April 2014, 09:45:55 UTC |
cd0c0de | Simon Willnauer | 09 April 2014, 09:37:02 UTC | [TEST] RecoveryWhileunderLoadTests sometimes need higher timeouts In some cases when we have a lot of docs with lots of shards recovery takes longer than 1m causing the tests to fail before all shards are recovered. This commit raises the timeout in this test to 5m max while it's rarely needed. This commit also adds an assertion to ElasticsearchAssertions that ensures that the cluster health requests are not hitting a timeout. | 09 April 2014, 09:37:02 UTC |
6d96683 | Boaz Leskes | 08 April 2014, 18:53:31 UTC | [Test] recoverWhileRelocating: Increase timeout while waiting for shards to replicate. | 08 April 2014, 18:53:31 UTC |
adc9a25 | Simon Willnauer | 08 April 2014, 12:20:34 UTC | Fix P/C assertions for rewrite reader Inner queries must be rewritten as soon as a weight is pulled ie. must be non-null. Closes #5731 | 08 April 2014, 15:47:31 UTC |
5b6fd6d | Simon Willnauer | 08 April 2014, 15:46:29 UTC | [TEST] Fix testRandomDirectoryIOExceptions to wait for green on reopen & add more assertions | 08 April 2014, 15:46:56 UTC |
a98b3fa | Simon Willnauer | 08 April 2014, 14:18:15 UTC | Revert "[TEST] Log where locks are created from if they are still open on close()" This reverts commit 7eb8b0d0c0b2ef3adcfd7b1e10ea0ca9a50c1adc. | 08 April 2014, 14:18:15 UTC |
7eb8b0d | Simon Willnauer | 08 April 2014, 13:14:08 UTC | [TEST] Log where locks are created from if they are still open on close() | 08 April 2014, 13:14:08 UTC |
de13d70 | Simon Willnauer | 08 April 2014, 10:40:37 UTC | [TEST] Wait for LANGUID events to be processed before pulling stats | 08 April 2014, 10:41:13 UTC |
e8467f0 | Boaz Leskes | 08 April 2014, 08:28:31 UTC | Failed shards could be re-assigned to the same nodes if multiple replicas failed at once After a shard fails on a node we assign a new replica on another node. This is important in order to avoid failing again due to node specific problems. In the rare case where two different replicas of the same shard failed in a short time span, we may fail to do so and assign one of them back to the node it's currently on. This happens if both shard failed events are processed within the same batch on the master. Closes #5725 | 08 April 2014, 10:10:45 UTC |
c58b823 | Simon Willnauer | 08 April 2014, 09:38:08 UTC | [TEST] Add more randomization to bulk tests | 08 April 2014, 09:38:08 UTC |
fd8a6ac | Simon Willnauer | 08 April 2014, 09:11:56 UTC | [TEST] make BulkTest more robust if test infra is slow | 08 April 2014, 09:11:56 UTC |
0cebbf1 | David Pilato | 08 April 2014, 08:55:26 UTC | Text Query has been replaced by Match Query Closes #5727. | 08 April 2014, 08:55:26 UTC |
b4c506b | Simon Willnauer | 08 April 2014, 08:07:53 UTC | [TEST] use no replicas in MLT tests - doc mappers need to be present on all nodes | 08 April 2014, 08:07:53 UTC |
dcc6a6e | Bill Hwang | 07 April 2014, 22:13:11 UTC | [BUILD] Remove site dependencies generation Removed dependencies library generation. It stalls Jenkins static analysis job | 07 April 2014, 23:01:02 UTC |
48031b6 | Andrew O'Brien | 07 April 2014, 21:47:39 UTC | Fixes typo in "Scan" search type documention | 07 April 2014, 22:01:37 UTC |
a1d0eee | Shay Banon | 07 April 2014, 20:34:47 UTC | [TEST] return the correct transport instance in mock transport | 07 April 2014, 20:34:56 UTC |
49c74e0 | Simon Willnauer | 07 April 2014, 20:09:09 UTC | Rename successulOps to successfulOps in TransportSearchTypeAction | 07 April 2014, 20:15:09 UTC |
befa833 | Simon Willnauer | 07 April 2014, 20:05:47 UTC | Make sure successful operations are correct if second phase is fast In TransportSearchTypeAction we need to increment successful ops first before we increment and compare the exit condition otherwise if we are fast we could concurrently update totalOps but then preempt one of the threads which can cause the successor to read a wrong value from successfulOps if second phase is very fast ie. searchType == count etc. This can cause wrong success stats in the search response. | 07 April 2014, 20:15:09 UTC |
f2181d5 | Simon Willnauer | 06 April 2014, 07:28:11 UTC | [TEST] Be more verbose if ClusterStatsTests fails | 07 April 2014, 20:14:10 UTC |
5138083 | Sean Gallagher | 01 April 2014, 16:28:08 UTC | Author: Sean Gallagher Date: Tue Apr 1 12:28:00 2014 Added upgrade.asciidoc and links to it from setup.asciidoc Author: Sean Gallagher Date: Apr 1 2014 Added upgrade.asciidoc Add upgrade instructions Author: Sean Gallagher Date: 4/4/14 Closes issue #5651 Fixed upgrade.asciidoc typo and incorrect usage. Author: Sean Gallagher Date: 4 Apr 2014 Closes 5651 | 07 April 2014, 18:43:35 UTC |