Commit Graph

5695 Commits

Author SHA1 Message Date
Clinton Gormley 94be785726 [DOCS] Added multi-index open/close 2013-09-04 23:20:55 +02:00
Clinton Gormley 6568dae12c [DOCS] Added "elastics" client for javascript 2013-09-04 23:20:55 +02:00
Clinton Gormley 5b60506b2e [DOCS] Added highlighting to the phrase suggester 2013-09-04 23:20:54 +02:00
Clinton Gormley 53ad7330fc [DOCS] Added docs for term vectors 2013-09-04 23:20:54 +02:00
Clinton Gormley eac2b3a52e [DOCS] Fixed typo 2013-09-04 23:20:54 +02:00
Simon Willnauer e68303d6a6 Allow test output to be configurable via CMD 2013-09-04 12:46:43 +02:00
Boaz Leskes f93efc3605 Added cluster state version to the debug logging of shards instances used in search. 2013-09-04 11:08:00 +02:00
Luca Cavanna e5b829303e Added client_ prefix to node clients created by TestCluster (client_node_#) 2013-09-04 10:37:00 +02:00
Luca Cavanna 74b00b63ea Node clients in tests use now the same naming convention as ordinary nodes (node#) 2013-09-04 10:26:18 +02:00
Simon Willnauer d5323231fe Clean up warnings in commons package. 2013-09-03 21:52:50 +02:00
Simon Willnauer 2ac7421f89 Remove UUID class in favor of Strings util methods 2013-09-03 21:52:50 +02:00
Clinton Gormley 393c28bee4 [DOCS] Removed outdated new/deprecated version notices 2013-09-03 21:28:31 +02:00
Martijn van Groningen d5a47e597d Removed ThreadLocals class and its references, since it is no longer needed. 2013-09-03 17:57:11 +02:00
Simon Willnauer b39961b2a6 Prevent ShardAllocator to modify the unassigned while running allocations
The unassinged list is used to make allocation decisions but is currently
modified during allocation runs which causes primaries to be throttled
during allocation. If this happens newly allocated indices can be stalled
for a long time turning a cluster into a RED state if concurrent relocations
and / or recoveries are happening.

Closes #3610
2013-09-03 17:09:44 +02:00
Luca Cavanna ca6c26c62d Improved test
Removed needless delete lines, added log lines and shared ensureGreen calls
2013-09-03 16:46:05 +02:00
Clinton Gormley 69d1d35fc1 [DOCS] Fixed an out of sequence header in the Groovy docs 2013-09-03 16:28:44 +02:00
Clinton Gormley 17234fe454 [DOCS] link: prefix not required when using {ref} attributes 2013-09-03 16:16:15 +02:00
Clinton Gormley e6127fc082 [DOCS] Chunk depth now configurable, so [float] not required 2013-09-03 16:15:50 +02:00
Luca Cavanna 2f10f94c13 Fixed PluginManagerTests to connect to the right node when checking if the plugin is available 2013-09-03 15:47:57 +02:00
Luca Cavanna 73a7b06d21 Switched PluginManagerTests to use default plugins location and added check that http connector is working properly 2013-09-03 15:08:43 +02:00
Luca Cavanna f9f664bc3b Added errorStream null check in HttpClient (used for testing purpose) 2013-09-03 14:18:04 +02:00
Shay Banon 058f57b198 test for primary allocating over replicas
still marked as await fix, requires investigation
2013-09-03 11:58:14 +02:00
Luca Cavanna 2d1b841ba1 Replaced localhost with 127.0.0.1 in PluginManagerTests 2013-09-03 11:44:56 +02:00
Luca Cavanna 45c8da3e98 Improved logic that removes top-level folder from archives when needed
Whether we remove the top-level folder from the archive depends now on the zip itself and not on where it was downloaded from. That makes it work installing local files too.

Closes #3582
2013-09-03 10:34:32 +02:00
Simon Willnauer eb2fed85f1 Add 'min_input_len' to completion suggester
Restrict the size of the input length to a reasonable size otherwise very
long strings can cause StackOverflowExceptions deep down in lucene land.
Yet, this is simply a saftly limit set to `50` UTF-16 codepoints by default.
This limit is only present at index time and not at query time. If prefix
completions > 50 UTF-16 codepoints are expected / desired this limit should be raised.
Critical string sizes are beyone the 1k UTF-16 Codepoints limit.

Closes #3596
2013-09-03 10:26:37 +02:00
Martijn van Groningen eed7f0bdb3 Restructured the transport action code for multi percolate api. 2013-09-03 10:19:58 +02:00
Boaz Leskes 7307e37efe Previously added logging in TransportRefreshAction caused NPE if search shards had errors (or was not yet started/assigned)
Better formatting for logging in TransportRefreshAction
2013-09-02 21:53:00 +02:00
Boaz Leskes a48c437ffb Search action now log in debug what shards/nodes were used before moving to second phase
Added  a debug log for executing a refresh request on a shard.
2013-09-02 17:17:53 +02:00
Luca Cavanna 91c3853452 Added new iteration to RecoveryWhileUnderLoadTests
We want to check whether the issue causing the test to fail would be fixed by an additional refresh or not
2013-09-02 16:29:49 +02:00
Luca Cavanna fc1d6670df Added log lines to log (debug level) acknowledgements regarding index alias creation and index state change 2013-09-02 12:16:14 +02:00
Luca Cavanna 38be478c46 Fixed close index when the index is already closed
Closes #3601
2013-09-02 10:58:09 +02:00
Luca Cavanna 206ab3a6d4 Used ElasticsearchAssertions to check count result in RecoveryWhileUnderLoadTests 2013-09-02 10:30:10 +02:00
Clinton Gormley ca4b85edef Added IDs to the community clients docs, to control HTML page names 2013-09-01 12:58:17 +02:00
Simon Willnauer 69848210d6 Allow up to 10% unbalance in random allocator tests 2013-08-31 09:13:54 +02:00
Boaz Leskes be09103258 SearchWhileCreatingIndexTests didn't always supply enough node for write consistency requirements and thus hanged.
Added more debug logging to TransportShardReplicationOperationAction

*Changed* node naming in tests from GUIDs to node# based naming as they are much easier to read
2013-08-30 20:56:23 +02:00
Martijn van Groningen be00437c65 Changed expectedOperationsPerItem from AtomicInteger array to AtomicReferenceArray<AtomicInteger>.
Made node stopping / starting less aggressive in RecoveryPercolatorTests.
2013-08-30 16:10:53 +02:00
Simon Willnauer c9c850167f Add more trace logging if shard allocation is throttled 2013-08-30 14:05:53 +02:00
Boaz Leskes c70b8fbd7b Added total, successful shard info to search related asserts in ElasticsearchAssertions
Rewrote SimpleSearchTests assertion to use that.
2013-08-30 14:02:13 +02:00
Simon Willnauer f8cf5ae7e9 Added Random test for ShardAllocator.
This test will make random allocation decision on a growing and
shrinking cluster leading to a random distribution of the shards.
After a certain amount of iterations the test allows allocation unless
the same shard is already allocated on a node and balances the cluster
to gain optimal balance.
2013-08-30 10:11:45 +02:00
Martijn van Groningen a7b2b7847a Use atomic collections to make sure all of the memory contents are visible from writing to reading thread. 2013-08-30 00:06:45 +02:00
Simon Willnauer 7113731022 Execute listeners on current thread if threadpool is shutting down 2013-08-29 16:40:38 +02:00
Simon Willnauer f408673bd4 Force default codec in tests where we rely on the latest lucene defaults 2013-08-29 16:40:38 +02:00
Nik Everett 10e55bd3ef Recheck cutoffScore during phrase_suggest merge.
The goal is to throw out suggestions that only meet the cutoff in some
shards.  This will happen if your input phrase is only contained in a
few shards.  If your shards are unbanced this rechecking can throw out
good suggestions.

Closes #3547.
2013-08-29 16:40:38 +02:00
Martijn van Groningen 76939b82d3 Removed unnecessary catch clause.
Improved logging
2013-08-29 16:00:59 +02:00
Boaz Leskes ffd019c07e Logging shard level multi percolate errors under debug rather than trace 2013-08-29 14:15:38 +02:00
Luca Cavanna 649108656d Added logging and asserts to make it easier to understand potential failures 2013-08-29 14:10:49 +02:00
Simon Willnauer 8b617fb48f Convert relative paths to URI first in tests to prevent URL encoded paths
If a path to the test classes contains spaces Hunspell tests failed due
to URL encoded paths. This happens on CI builds if you give the build
a name containing a space. This is fixed by first converting to a URI
and created a File object from the URI directly.
2013-08-29 12:56:33 +02:00
Martijn van Groningen 25c1b93d57 Ignore AlreadyExpiredException exception during when indexing a document on a replica.
The document should just be indexed to stay consistent with the primary shard.
2013-08-29 11:38:13 +02:00
Martijn van Groningen a1f2f44eef Make sure 2 writes happened, so we can check if 2 deletes happen that have been triggered by ttl 2013-08-29 11:27:01 +02:00
Adrien Grand 5b6be0c456 Use a separate build directory for Eclipse.
The fact that Maven and Eclipse share the same build directories can trigger
race conditions when both are trying to build at the same time, eg. if you run
`mvn clean test` while Eclipse is up and running: Eclipse will notice that some
class files are missing and start compiling in parallel with Maven.
2013-08-29 10:29:26 +02:00