Commit Graph

10251 Commits

Author SHA1 Message Date
Simon Willnauer 8736543c71 [RECOVERY] Ensure shards are identical after recovery
Today we don't check if the recovery target has all the
files that we expect there after the recovery. This commit
adds aditional safety to ensure all files are present with the
correct checksums on recovery finalization.

Closes #8723
2014-12-02 14:05:44 +01:00
Simon Willnauer c1edcaf388 [RECOVERY] Make recovery delay configurable
Today we wait 500ms before we retry a recovery if the target node is not ready.
This happens if the source starts the recovery before the target has
processed the clusterstate moving the target shard into the right state.
This can cause a 500ms delay each time it happens while the shard is ready
way earlier on the target node. This commit makes this delay configurable
to mainly speed up test processing and shard allocation in tests.
2014-12-02 13:33:35 +01:00
John Michael Luy f20f6ffe22 Docs: Update range-query.asciidoc
Closes #8740
2014-12-02 12:55:44 +01:00
Martijn van Groningen d7e224da04 Added `inner_hits` feature that allows to include nested hits.
Inner hits allows to embed nested inner objects, children documents or the parent document that contributed to the matching of the returned search hit as inner hits, which would otherwise be hidden.

Closes #8153
Closes #3022
Closes #3152
2014-12-02 12:01:01 +01:00
Thilo Fromm e92ff00192 bin/elasticsearch: add help, fix endless loop
This change adds command line help for all options to the es start script.
Both '-h' and '--help' options are accepted.

Also, an endless busy loop in the long options parser was fixed: running the
script with a long opt parameter w/o value (e.g. "elasticsearch --buuuurrrnn")
the long option parser would end up in an endless busy loop.

Signed-off-by: Thilo Fromm <github@thilo-fromm.de>
2014-12-02 11:19:46 +01:00
Simon Willnauer 942e752ac1 Remove unused member / argument on Store 2014-12-02 10:11:34 +01:00
Simon Willnauer 9b5b281fe8 [TEST] Ensure we have a mapping for all types on open/close 2014-12-02 10:11:04 +01:00
David Pilato 317192b647 java: QueryBuilders cleanup (add and deprecate)
Some QueryBuilders are missing or have a different naming than the other ones.

This patch is applied to branch 1.x and master (elasticsearch 1.5 and 2.0):

Added
-----

* `templateQuery(...)`
* `commonTermsQuery(...)`
* `queryStringQuery(...)`
* `simpleQueryStringQuery(...)`

Deprecated
----------

* `commonTerms(...)`
* `queryString(...)`
* `simpleQueryString(...)`
2014-12-01 14:41:25 +01:00
David Pilato 45c24b2eb3 Remove java facet documentation in master
As we did in reference guide, we need to remove facet java code in master branch.
2014-12-01 14:22:48 +01:00
Itamar Syn-Hershko cb042cd662 Fixing typo
Closes #8713
2014-12-01 10:52:00 +01:00
Simon Willnauer ade5aaae5f [TEST] Ensure that all flushes happen on PeerRecovery tests 2014-11-30 16:31:49 +01:00
Simon Willnauer c630b1e8a4 [TEST] Move NoMergePolicyProvider into it's own class 2014-11-30 16:31:49 +01:00
Clinton Gormley c4fd1247c0 Docs: Fixed bad link to top-hits agg 2014-11-30 09:38:57 +01:00
Simon Willnauer ca9abb1caf [TEST] wait for all shards to be allocated before IndexStatsTests runs 2014-11-29 20:11:21 +01:00
Simon Willnauer 539faf4e65 [TEST] Mute CircuitBreakerServiceTests.testMemoryBreaker
Relates to #8710
2014-11-29 20:07:19 +01:00
David Pilato d7d937300a Add java documentation for aggregations
Buckets:
--------

* terms
* range
* global
* filter
* filters
* missing
* nested
* reverse nested
* children
* significant terms
* date range
* ip range
* range
* histogram
* date histogram
* geo distance
* geo hash grid

Metrics:
--------

* min
* max
* sum
* avg
* stats
* extended stats
* value count
* percentiles
* percentile rank
* cardinality
* geo bounds
* top hits
* scripted metric
2014-11-29 19:46:33 +01:00
Simon Willnauer ef8802d878 [TEST] make sure number of shard is low in network corruption tests 2014-11-29 17:16:46 +01:00
Simon Willnauer 2d0309f0d4 [TEST] Use private random instance to build test cluster config 2014-11-29 16:06:05 +01:00
Simon Willnauer 75cc8ee097 [TEST] Speed up recoveries if tests.nighly=true 2014-11-29 15:37:15 +01:00
Simon Willnauer 7d3da915b0 [TEST] Don't fail test if dummy doc is not found
Relates to #8706
2014-11-29 15:36:05 +01:00
Dan Tuffery 3b5fa9075a Docs: Grammar correction
Closes #8702
2014-11-29 14:06:04 +01:00
Lee Hinman 7776b6b4f0 [TEST] Use a pooling connection manager for REST tests 2014-11-28 22:04:01 +01:00
Simon Willnauer 29422c645b [TEST] Try to speed up REST tests by reducing max number of replicas and shards 2014-11-28 21:16:38 +01:00
Patrick Koenig 76157b8b90 Logging - provides template usage information on index creation
Closes #7421
2014-11-28 16:11:16 +00:00
Clinton Gormley 88e06cba80 Update daterange-aggregation.asciidoc
Clarified the date-math expressions on date range aggregations

Closes #8703
2014-11-28 16:53:33 +01:00
Lee Hinman 1d8fd0fc04 [TEST] Explicit wait for fielddata breaker to be cleared 2014-11-28 15:56:11 +01:00
Lee Hinman f9d7e76928 [TEST] Relax constraints of breaker tests even more
They were very stringent, and add values in worst-case situations. The
new values are acceptable while still testing the functionality of the
breaker.
2014-11-28 15:12:09 +01:00
Alex Ksikes 256712640f MLT Query: Support for ignore docs
Adds a `ignore_like` parameter to the MLT Query, which simply tells the
algorithm to skip all the terms from the given documents. This could be useful
in order to better guide nearest neighbor search by telling the algorithm to
never explore the space spanned by the given `ignore_like` docs. In essence we
are interested about the characteristic of a given item, but not of the ones
provided by `ignore_like`, thereby forcing the algorithm to go deeper in its
selection of terms. Note that this is different than simply performing a must
not boolean query on the unliked items. The syntax is exactly the same as the
`like` parameter.

Closes #8674
2014-11-28 14:48:43 +01:00
pmamat 9e2eaeece4 Docs: Additional info about _score calculation
Description taken from http://www.elasticsearch.org/guide/en/elasticsearch/guide/current/multi-query-strings.html / 110_Multi_Field_Search/05_Multiple_query_strings.asciidoc

Closes #8635
2014-11-28 13:54:45 +01:00
Britta Weber 59507cf793 function_score: match only document with score above custom score threshold
functon_score matched each document regardless of the computed score.
This commit adds a query parameter `min_score` (-Float.MAX_VALUE default).
Documents that have a score lower than this threshold will not be mached.

closes #6952
2014-11-28 12:35:26 +01:00
Simon Willnauer 93b52c925d [TEST] With pipelining disabled requests can come back in any order
Closes #8697
2014-11-28 12:28:31 +01:00
Simon Willnauer 5c6c7f23ba [TEST] add back accidetially removed test logging 2014-11-28 11:01:15 +01:00
Simon Willnauer bc563931c3 [TEST] move test to a different class that can disable publish timeout on demand 2014-11-28 10:49:56 +01:00
Simon Willnauer c524e469ec [TEST] remove outdated TestLogging annotation
Conflicts:
	src/test/java/org/elasticsearch/bwcompat/UnicastBackwardsCompatibilityTest.java
	src/test/java/org/elasticsearch/recovery/RelocationTests.java
2014-11-28 10:44:06 +01:00
Lee Hinman 600f02b407 [TEST] Add an assert for null indices in InternalEngineIntegrationTest 2014-11-28 10:29:24 +01:00
Boaz Leskes 8456489773 Test: add trace logging to RelocationTests
also improved error message when failing to delete a dummy doc
2014-11-28 10:22:49 +01:00
Simon Willnauer b18675efb4 [TEST] Use different index on each iteration to ensure less timeout prone tests 2014-11-28 10:10:10 +01:00
Alex Ksikes d7338ffdbc MLT Query: Fix exclude with artificial documents
Artificial documents get assigned a random id. When include is set to false
(default), the ids of these documents also get included, when they should
rather be ignored.

Closes #8679
2014-11-28 08:07:57 +01:00
Simon Willnauer fe762c0eb5 [TEST] Reduce possible number of indices in the test - 10 indices can create tons of shards 2014-11-27 23:04:40 +01:00
David Pilato 43a1435d3b [Docs] fix consistency between examples 2014-11-27 20:29:34 +01:00
David Pilato 40f0e07db3 [Docs] Fix missing new line 2014-11-27 19:39:12 +01:00
Britta Weber f00b431c18 [docs] explain default settings for parameters of decay functions
relates to #8624
2014-11-27 19:18:55 +01:00
Martijn van Groningen 06c39e79d6 Test: predefine sort fields in mapping, otherwise during the test the field may not be found if it were to be introduced dynamically at index time. 2014-11-27 17:57:35 +01:00
Simon Willnauer 6e4f3cf42d [CLIENT] Only fetch the node info during node sampling
Today we are fetching a lot of information that is unneeded
for the sampling phase. We only really need the DiscoveryNode
to ensure the node is still there.
This commit clears all flags to be false on the NodeInfo call.
2014-11-27 17:12:38 +01:00
Simon Willnauer c1832e36dc [HEALTH] Don't mark health as timed out if desired state is reached
Today we mark cluster health requests as timed out if the request has already
timed out. Yet, and implementation detail of the health request is that we are
waiting for events which can take quite some time if the machine is busy. If
we have already reached a valid state while waiting for events to be processed
we shouldn't mark the clusterstate as timed-out. This will help tests that
wait for green state with lots of nodes and shards to not fail the cluster health
calls.
2014-11-27 17:11:07 +01:00
Boaz Leskes cd717ab8e7 Test: UpdateTests.stressUpdateDeleteConcurrency shouldn't turn off threaded operations
This may result in all network threads being busy for too long.
2014-11-27 16:30:26 +01:00
javanna 0c70b3ca7d [TEST] remove duplicated tests.slow from randomized testing maven plugin conf 2014-11-27 11:36:15 +01:00
David Pilato da27c2104a [Docs] Fix missing comma in mapping 2014-11-27 11:03:19 +01:00
javanna bf0387e0bc [TEST] make sure rest tests info is printed for any @Rest annotated test
We introduced the @Rest annotation a while ago for REST tests (see #7795), we have then to make sure that relevant info to reproduce failures gets printed out for any test that is marked with such annotation, not only for ElasticsearchRestTests

Closes #8680
2014-11-27 10:42:25 +01:00
javanna e07b0deecd [TEST] Extend unicast ports generation to support more concurrent clusters
Make it possible to run multiple tests with unicast configuration, by assigning ports based on their test scope.
Every jvm still gets its own port range based on the jvm id, but we now make sure that the different jvms ranges never overlap. The global cluster gets a reserved port range, while SUITE and TEST scopes are treated equally, just assuming that they never run concurrently on the same jvm, thus ports can be safely reused.

Closes #8634
2014-11-27 09:02:23 +01:00