Commit Graph

570 Commits

Author SHA1 Message Date
Nishant 6099d20303 [FIX] ReleaseException when the path is being written by multiple tasks (#3494)
* fix ReleaseException when the path is being written by multiple task

* Do not throw IOException if another replica wins the race for segment creation

fix if check

* handle logging comments

* fix test
2016-09-22 14:25:41 -05:00
Navis Ryu 74e1243c7e Fix test fail of PollingLookupTest.testApplyAfterDataChange (#3489) 2016-09-22 08:33:59 -07:00
Himanshu 05ea88df5c fix kafka-indexing-service pom to not reference specific version but parent version for druid core dependencies (#3472) 2016-09-20 15:18:21 -07:00
David Lim 96fcca18ea update KafkaSupervisor to make HTTP requests to tasks in parallel where possible (#3452) 2016-09-20 22:51:15 +05:30
Slim 3175e17a3b Cached lookup module. first cut implementing JDBC cache (#2819) 2016-09-16 13:45:54 -07:00
Charles Allen 95e08b38ea [QTL] Reduced Locking Lookups (#3071)
* Lockless lookups

* Fix compile problem

* Make stack trace throw instead

* Remove non-germane change

* * Add better naming to cache keys. Makes logging nicer
* Fix #3459

* Move start/stop lock to non-interruptable for readability purposes
2016-09-16 11:54:23 -07:00
Gleb Smirnov d981a2aa02 Avoid interrupting ZookeeperConsumerConnector.shutdown() #3346 (#3403) 2016-09-14 17:44:27 -07:00
Himanshu a069257d37 avro-extension -- feature to specify multiple avro reader schemas inline (#3368)
* rename SimpleAvroBytesDecoder to InlineSchemaAvroBytesDecoder

* feature to specify multiple schemas inline in avro module
2016-09-13 14:54:31 -07:00
Gian Merlino bcff08826b KafkaIndexTask: Treat null values as unparseable. (#3453) 2016-09-13 10:56:38 -07:00
Slim ba6ddf307e Adding hadoop kerberos authentification. (#3419)
* adding kerberos authentication

* make the 2 functions identical
2016-09-13 10:42:50 -07:00
Jonathan Wei df766b2bbd Add dimension handling interface for ingestion and segment creation (#3217)
* Add dimension handling interface for ingestion and segment creation

* update javadocs for DimensionHandler/DimensionIndexer

* Move IndexIO row validation into DimensionHandler

* Fix null column skipping in mergerV9

* Add deprecation note for 'numeric_dims' filename pattern in IndexIO v8->v9 conversion

* Fix java7 test failure
2016-09-12 12:54:02 -07:00
Alexander Saydakov 1a5042ca26 updated dependency on sketches-core (#3443)
* updated dependency on sketches-core to 0.7.0

* Use sketches-core-0.4.1, which is the latest version still compatible
with JDK7
2016-09-09 16:21:32 -07:00
David Lim 146a17de48 KafkaIndexTask: allow pause to break out of retry loop (#3401) 2016-09-06 22:29:37 -06:00
David Lim 5b1ae21bd1 retry calls to getStartTime (#3429) 2016-09-06 14:02:22 -07:00
Stéphane Derosiaux 48dce88aab Add flag binaryAsString for parquet ingestion (#3381) 2016-08-30 17:30:50 -07:00
David Lim ed924bf214 allow registrants to opt out of announcing themselves when registering as a chat handler (#3360) 2016-08-16 10:51:28 +05:30
Himanshu 70d99fe3c6 Initialize ApproximateHistogram Module in ApproximateHistogramGroupByQueryTest (#3363)
or else the test fails if ran independently.
2016-08-15 10:19:33 -07:00
Himanshu 46da682231 avro-extensions -- feature to specify avro reader schema inline in the task json for all events (#3249) 2016-08-10 10:49:26 -07:00
Jonathan Wei 890e3bdd3f More informative query unit test names (#3342) 2016-08-09 22:24:48 -07:00
Jonathan Wei decefb7477 Add time interval dim filter and retention analysis example (#3315)
* Add time interval dim filter and retention analysis example

* Use closed-open matching for intervals, update cache key generation

* Fix time filtering tests for interval boundary change
2016-08-05 07:25:04 -07:00
Navis Ryu 5b3f0ccb1f Support variance and standard deviation (#2525)
* Support variance and standard deviation

* addressed comments
2016-08-04 17:32:58 -07:00
Gleb Smirnov 33dbe0800c Makes kafka lookup extraction factory's replace() behavior consistent with other lookup extraction factories (#3326) 2016-08-04 10:24:19 -07:00
Gian Merlino 8030f1cb67 Be more respectful of maxRowsInMemory. (#3284)
- Appenderator: Respect maxRowsInMemory across all sinks.
- KafkaIndexTask: Respect maxRowsInMemory across all partitions.
2016-07-26 15:02:35 -06:00
Charles Allen 3f1681c16c Caffeine cache extension (#3028)
* Initial commit of caffeine cache

* Address code comments

* Move and fixup README.md a bit

* Improve caffeine readme information

* Cleanup caffeine pom

* Address review comments

* Bump caffeine to 2.3.1

* Bump druid version to 0.9.2-SNAPSHOT

* Make test not fail randomly.

See https://github.com/ben-manes/caffeine/pull/93#issuecomment-227617998 for an explanation

* Fix distribution and documentation

* Add caffeine to extensions.md

* Fix links in extensions.md

* Lexicographic
2016-07-06 15:42:54 -07:00
Charles Allen bfa5c05aaa Make global lookup cache introspector class public (#3199)
* Make global lookup cache introspector class public
* Fixes #3187

* Make KafkaLookupExtractorIntrospectionHandler a public static class
2016-07-01 15:50:57 -07:00
Xavier Léauté 485e381387 remove datasource from hadoop output path (#3196)
fixes #2083, follow-up to #1702
2016-06-29 08:53:45 -07:00
David Lim 1d40df4bb7 fix kafka consumer concurrent access during shutdown (#3193) 2016-06-28 13:23:17 -07:00
Hyukjin Kwon 45f553fc28 Replace the deprecated usage of NoneShardSpec (#3166) 2016-06-25 10:27:25 -07:00
Gian Merlino 4cc39b2ee7 Alternative groupBy strategy. (#2998)
This patch introduces a GroupByStrategy concept and two strategies: "v1"
is the current groupBy strategy and "v2" is a new one. It also introduces
a merge buffers concept in DruidProcessingModule, to try to better
manage memory used for merging.

Both of these are described in more detail in #2987.

There are two goals of this patch:

1. Make it possible for historical/realtime nodes to return larger groupBy
   result sets, faster, with better memory management.
2. Make it possible for brokers to merge streams when there are no order-by
   columns, avoiding materialization.

This patch does not do anything to help with memory management on the broker
when there are order-by columns or when there are nested queries. That could
potentially be done in a future patch.
2016-06-24 18:06:09 -07:00
du00cs ebd654228b fix: avro types exception in sketch (#3167) 2016-06-22 15:54:20 -05:00
Charles Allen 674f94083e Add more logging around failed S3DataSegmentMover DeleteExceptions (#3104)
* Add more logging around failed S3DataSegmentMover DeleteExceptions

* Fix test NPE
2016-06-16 14:58:33 -07:00
Charles Allen f7fa1d8c62 [QTL] Allow S3 version finder to search entire s3 object key (#3139)
* Allow S3 version finder to search entire s3 object key
* Previously only was able to search immediate "directory"

* Update method javadoc

* Expand docs a bit better
2016-06-13 21:02:28 -07:00
Gian Merlino ebf890fe79 Update master version to 0.9.2-SNAPSHOT. (#3133) 2016-06-13 13:10:38 -07:00
David Lim 4faa298977 update kafka client for kafka indexing service to 0.9.0.1 (#3109) 2016-06-08 06:51:03 -07:00
Charles Allen 8cac710546 Async lookups-cached-global by default (#3074)
* Async lookups-cached-global by default
* Also better lookup docs

* Fix test timeouts

* Fix timing of deserialized test

* Fix problem with 0 wait failing immediately
2016-06-03 15:58:10 -05:00
David Lim a2290a8f05 support seamless config changes (#3051) 2016-06-03 13:50:19 -07:00
Charles Allen 447033985e Make S3DataSegmentMover not bother checking for items if they are the same (#3032)
* Make S3DataSegmentMover not bother checking for items if they are the same
2016-06-02 17:27:21 +01:00
David Lim f6c39cc844 Kafka task minimum message time (#3035)
* add KafkaIndexTask support for minimumMessageTime

* add Kafka supervisor support for lateMessageRejectionPeriod
2016-05-31 11:37:00 -07:00
David Lim 3ef24c03b3 Validate X-Druid-Task-Id header in request/response and support retrying on outdated TaskLocation information, add KafkaIndexTaskClient unit tests (#3006)
* validate X-Druid-Task-Id header in request and add header to response

* modify KafkaIndexTaskClient to take a TaskLocationProvider as the TaskLocation may not remain constant
2016-05-25 22:05:18 -07:00
Charles Allen 8024b915e2 [QTL] Implement LookupExtractorFactory of namespaced lookup (#2926)
* support LookupReferencesManager registration of namespaced lookup and eliminate static configurations for lookup from namespecd lookup extensions

- druid-namespace-lookup and druid-kafka-extraction-namespace are modified
- However, druid-namespace-lookup still has configuration about ON/OFF
  HEAP cache manager selection, which is not namespace wide
  configuration but node wide configuration as multiple namespace shares
  the same cache manager

* update KafkaExtractionNamespaceTest to reflect argument signature changes

* Add more synchronization functionality to NamespaceLookupExtractorFactory

* Remove old way of using extraction namespaces

* resolve compile error by supporting LookupIntrospectHandler

* Remove kafka lookups

* Remove unused stuff

* Fix start and stop behavior to be consistent with new javadocs

* Remove unused strings

* Add timeout option

* Address comments on configurations and improve docs

* Add more options and update hash key and replaces

* Move monitoring to the overriding classes

* Add better start/stop logging

* Remove old docs about namespace names

* Fix bad comma

* Add `@JsonIgnore` to lookup factory

* Address code review comments

* Remove ExtractionNamespace from module json registration

* Fix problems with naming and initialization. Add tests

* Optimize imports / reformat

* Fix future not being properly cancelled on failed initial scheduling

* Fix delete returns

* Add more docs about whole introspection

* Add `/version` introspection point for lookups

* Add more tests and address comments

* Add StaticMap extraction namespace for testing. Also add a bunch of tests

* Move cache system property to `druid.lookup.namespace.cache.type`

* Make VERSION lower case

* Change poll period to 0ms  for StaticMap

* Move cache key to bytebuffer

* Change hashCode and equals on static map extraction fn

* Add more comments on StaticMap

* Address comments

* Make scheduleAndWait use a latch

* Sanity renames and fix imports

* Remove extra info in docs

* Fix review comments

* Strengthen failure on start from warn to error

* Address comments

* Rename namespace-lookup to lookups-cached-global

* Fix injective mis-naming
* Also add serde test
2016-05-24 10:56:40 -07:00
Charles Allen 15ccf451f9 Move QueryGranularity static fields to QueryGranularities (#2980)
* Move QueryGranularity static fields to QueryGranularityUtil
* Fixes #2979

* Add test showing #2979

* change name to QueryGranularities
2016-05-17 16:23:48 -07:00
Himanshu d3e9c47a5f use correct ObjectMapper in Index[IO/Merger] in AggregationTestHelper and minor fix in theta sketch SketchMergeAggregatorFactory.getMergingFactory(..) (#2943) 2016-05-13 10:06:31 +05:30
Slim 45b2e65d75 [QTL] adding listDelimiter to lookup parser spec (#2941)
* adding listDelimiter to lookup parser spec

* cleaning code
2016-05-10 15:41:16 +05:30
Charles Allen 90b0b0a4ad Make URIExtraction not require FileSystem impls for URIs it understands (#2929)
* Make URIExtraction not require FileSystem impls for URIs it understands
* Fixes #2928

* Preserve URI information

* Simply case for exact matching

* Move unused variable
2016-05-08 23:23:53 +05:30
David Lim b489f63698 Supervisor for KafkaIndexTask (#2656)
* supervisor for kafka indexing tasks

* cr changes
2016-05-04 23:13:13 -07:00
Charles Allen 2a769a9fb7 Make S3DataSegmentPuller do GET requests less often (#2900)
* Make S3DataSegmentPuller do GET requests less often
* Fixes #2894

* Run intellij formatting on S3Utils

* Remove forced stream fetching on getVersion

* Remove unneeded finalize

* Allow initial object fetching to fail and be retried
2016-05-04 16:21:35 -07:00
Gian Merlino f8ddfb9a4b Split SegmentInsertAction and SegmentTransactionalInsertAction for backwards compat. (#2922)
Fixes #2912.
2016-05-04 13:54:34 -07:00
Charles Allen 6b957aa072 [QTL] Make URI Exctraction Namespace take more sane arguments (#2738)
* Make URI Exctraction Namespace take more sane arguments
* Fixes https://github.com/druid-io/druid/issues/2669

* Update docs

* Rename error message

* Undo overzealous deletion of docs

* Explain caching mechanism a bit more in docs
2016-05-02 12:54:34 -07:00
Charles Allen 54b717bdc3 [QTL] Move kafka-extraction-namespace to the Lookup framework. (#2800)
* Move kafka-extraction-namespace to the Lookup framework.

* Address comments

* Fix missing kafka introspection

* Fix tests to be less racy

* Make testing a bit more leniant

* Make tests even more forgiving

* Add comments to kafka lookup cache method

* Move startStopLock to just use started

* Make start() and stop() idempotent

* Forgot to update test after last change, test now accounts for idempotency

* Add extra idempotency on stop check

* Add more descriptive docs of behavior
2016-05-02 09:45:13 -07:00
Gian Merlino 67b47c982f Datasketches: Remove isInputThetaSketch from cache key. (#2899) 2016-04-28 18:14:52 -07:00
Gian Merlino 16080dc54f Adjust colliding aggregator cache IDs. (#2891)
- Renumbered ApproximateHistogramAggregatorFactory from 8 to 12,
  8 was taken by CardinalityAggregatorFactory
- Renumbered ApproximateHistogramFoldingAggregatorFactory from 9 to 13,
  9 was taken by FilteredAggregatorFactory
2016-04-28 10:11:33 -07:00
Gian Merlino 909abd17f3 Sketch cache key should include size, isInputThetaSketch. (#2893) 2016-04-28 10:10:46 -07:00
David Lim 7641f2628f add control and status endpoints to KafkaIndexTask (#2730) 2016-04-21 15:34:59 -07:00
Xavier Léauté 5938d9085b Stream segments from database (#2859)
* Avoids fetching all segment records into heap by JDBC driver
* Set connection to read-only to help database optimize queries
* Update JDBC drivers (MySQL has fixes for streaming results)
2016-04-21 05:40:07 +08:00
Gian Merlino 08c784fbf6 KafkaIndexTask: Use a separate sequence per Kafka partition in order to make (#2844)
segment creation deterministic.

This means that each segment will contain data from just one Kafka
partition. So, users will probably not want to have a super high number
of Kafka partitions...

Fixes #2703.
2016-04-18 22:29:52 -07:00
Xavier Léauté 0f8a037bcd support PostgreSQL >= 9.5 upsert capability 2016-04-01 16:53:27 -07:00
Gian Merlino 977e867ad8 Downgrade geoip2, exclude com.google.http-client.
Reverts "Update com.maxmind.geoip2 to 2.6.0" and exclude the google http client
from com.maxmind.geoip2. This should satisfy the original need from #2646 (wanting
to run Druid along with an upgraded com.google.http-client) while preventing
Jackson conflicts pointed out in #2717.

Fixes #2717.

This reverts commit 21b7572533.
2016-03-25 14:43:22 -07:00
Himanshu f26e73d133 Merge pull request #2720 from gianm/druid-api
Move druid-api into the druid repo.
2016-03-24 15:51:10 -05:00
Gian Merlino 7e7a886f65 Move druid-api into the druid repo.
This is from druid-api-0.3.17, as of commit 51884f1d05d5512cacaf62cedfbb28c6ab2535cf
in the druid-api repo.
2016-03-24 11:04:34 -07:00
Himanshu Gupta 4aead38130 fix SketchEstimate post aggregator's getComparator() and test changes to verify same 2016-03-24 10:11:06 -05:00
jon-wei a59c9ee1b1 Support use of DimensionSchema class in DimensionsSpec 2016-03-21 13:12:04 -07:00
Gian Merlino 738dcd8cd9 Update version to 0.9.1-SNAPSHOT.
Fixes #2462
2016-03-17 10:34:20 -07:00
Slim cf342d8d3c Merge pull request #2517 from b-slim/adding_lookup_snapshot_utility
[QTL][Lookup] lookup module with the snapshot utility
2016-03-17 11:39:47 -05:00
Slim Bouguerra 0c86b29ef0 lookup module with the snapshot utility 2016-03-17 09:20:41 -05:00
Charles Allen 02805a74a1 Merge pull request #2648 from chtefi/master
Ignore case when testing for table existence
2016-03-14 13:57:53 -07:00
Stéphane Derosiaux 416cb03687 Ignore case when testing for table existence 2016-03-13 11:17:30 +01:00
Gian Merlino f22fb2c2cf KafkaIndexTask.
Reads a specific offset range from specific partitions, and can use dataSource metadata
transactions to guarantee exactly-once ingestion.

Each task has a finite lifecycle, so it is expected that some process will be supervising
existing tasks and creating new ones when needed.
2016-03-10 18:41:43 -08:00
Gian Merlino 187569e702 DataSource metadata.
Geared towards supporting transactional inserts of new segments. This involves an
interface "DataSourceMetadata" that allows combining of partially specified metadata
(useful for partitioned ingestion).

DataSource metadata is stored in a new "dataSource" table.
2016-03-10 17:41:50 -08:00
Nishant ba1185963b Fix a bunch of dependencies
* Eliminate exclusion groups from pull-deps
* Only consider dependency nodes in pull-deps if they are not in the following scopes
	* provided
	* test
	* system
* Fix a bunch of `<scope>provided</scope>` missing tags
* Better exclusions for a couple of problematic libs
2016-03-10 10:18:08 -08:00
fjy e3e932a4d4 refactor extensions into core and contrib 2016-03-08 17:12:09 -08:00