8e8d454 | Ufuk Celebi | 10 October 2016, 13:26:32 UTC | Commit for release 1.1.3 | 10 October 2016, 13:26:32 UTC |
d619f51 | Stephan Ewen | 10 October 2016, 12:33:57 UTC | [FLINK-4788] [streaming api] Fix state backend classloading from configuration | 10 October 2016, 12:40:52 UTC |
7267562 | Maximilian Michels | 07 October 2016, 18:06:18 UTC | [FLINK-4777] catch IOException in ContinuousFileMonitoringFunction FileSystem.listStatus(path) may throw an IOException when it lists files and then retrieves their file status. This is quite common, e.g. editors which create temporary files and move them. The ContinuousFileMonitoringFunction can only apply a file path filter afterwards. The solution is to defer file checks until no exception is caught anymore. This closes #2610. | 10 October 2016, 08:46:41 UTC |
bab59df | kl0u | 04 October 2016, 13:27:59 UTC | [FLINK-4329] Fix Streaming File Source Timestamps/Watermarks Handling This closes #2593. | 05 October 2016, 15:52:26 UTC |
fddd89b | Maximilian Michels | 30 September 2016, 13:00:32 UTC | [FLINK-4677] fail if user jar contains no executions This closes #2548. | 05 October 2016, 14:09:43 UTC |
a83dbae | Tzu-Li (Gordon) Tai | 05 October 2016, 10:17:34 UTC | [hotfix] [kafka] Committed offset value set in KafkaTopicPartitionState should also be incremented by 1. The broken behaviour was introduced in the last hotfix commit eece0dd0. | 05 October 2016, 10:50:14 UTC |
a31a22e | Maximilian Michels | 04 October 2016, 09:12:35 UTC | [FLINK-4732] remove maven junction plugin On Windows, the plugin downloads and executes code from the author's web site. The downloaded file is not signed in the same way as Maven artifacts from Maven central which have to be signed with the developer's key. This could be a potential target for attackers. This closes #2586 | 04 October 2016, 13:49:09 UTC |
98b399d | Niels Basjes | 04 October 2016, 10:13:12 UTC | [FLINK-4311] [hbase-connector] Fixed TableInputFormat to process multiple input splits. | 04 October 2016, 12:43:53 UTC |
6f33361 | Fabian Hueske | 04 October 2016, 12:43:35 UTC | [hotfix] [kafka] Fix NPE in Kafka09Fetcher | 04 October 2016, 12:43:52 UTC |
400c49c | Tzu-Li (Gordon) Tai | 01 October 2016, 09:20:19 UTC | [FLINK-4618] [kafka-connector] Minor improvements to comment and variable naming This closes #2579 | 01 October 2016, 10:16:51 UTC |
bef0ebe | Max Kuklinski | 30 September 2016, 21:03:30 UTC | [FLINK-4618] [kafka-connector] Incremented the commited offset by one to avoid duplicate read message. | 01 October 2016, 10:16:35 UTC |
90d7759 | Stephan Ewen | 29 September 2016, 16:09:51 UTC | [FLINK-4702] [kafka connector] Commit offsets to Kafka asynchronously and don't block on polls Letting the Kafka commit block on polls means that 'notifyCheckpointComplete()' may take very long. This is mostly relevant for low-throughput Kafka topics. | 30 September 2016, 10:39:53 UTC |
caa0fbb | Stephan Ewen | 23 September 2016, 16:42:47 UTC | [FLINK-4672] [taskmanager] Do not decorate Actor Kill messages | 26 September 2016, 09:25:01 UTC |
62c666f | Maximilian Michels | 14 September 2016, 09:00:58 UTC | [FLINK-4485] close and remove user class loader after job completion Keeping the user class loader around after job completion may lead to excessive temp space usage because all user jars are kept until the class loader is garbage collected. Tests showed that garbage collection can be delayed for a long time after the class loader is not referenced anymore. Note that for the class loader to not be referenced anymore, its job has to be removed from the archive. The fastest way to minimize temp space usage is to close and remove the URLClassloader after job completion. This requires us to keep a serializable copy of all data which needs the user class loader after job completion, e.g. to display data on the web interface. | 24 September 2016, 12:21:07 UTC |
8fd08bf | swapnil-chougule | 22 September 2016, 12:31:20 UTC | [FLINK-4663] [jdbc] Fix JDBCOutputFormat log message This closes #2534 | 23 September 2016, 12:16:01 UTC |
52a4440 | Stephan Ewen | 20 September 2016, 18:49:40 UTC | [FLINK-4640] [streaming api] Ensure that the state descriptors properly initialize the serializers. | 21 September 2016, 13:12:02 UTC |
a7f6594 | Fabian Hueske | 16 September 2016, 16:40:32 UTC | [FLINK-2662] [dataSet] [optimizer] Fix merging of unions with multiple outputs. Translate union with N outputs into N unions with single output. | 20 September 2016, 19:55:52 UTC |
c5b391c | Aljoscha Krettek | 07 September 2016, 11:51:53 UTC | [FLINK-4589] [DataStream API] Fix Merging of Covering Window in MergingWindowSet This also adds two new test cases for that problem. This closes #2476 | 15 September 2016, 17:32:42 UTC |
f263b99 | Stephan Ewen | 02 September 2016, 09:45:25 UTC | [FLINK-4566] [network runtime] Properly preserve exception causes for ProducerFailedException | 02 September 2016, 15:41:41 UTC |
e5b4f46 | Stephan Ewen | 02 September 2016, 09:38:53 UTC | [FLINK-4567] [runtime] Enhance SerializedThrowable to properly mimic Exception causes | 02 September 2016, 15:41:33 UTC |
df72667 | Tzu-Li (Gordon) Tai | 29 August 2016, 09:30:39 UTC | [FLINK-4514][kinesis-connector] Handle unexpected ExpiredIteratorExceptions This closes #2432 | 30 August 2016, 15:55:47 UTC |
0be5581 | Robert Metzger | 30 August 2016, 14:56:41 UTC | [hotfix] Set the quickstart version correctly | 30 August 2016, 14:56:45 UTC |
f7ec1ef | Maximilian Michels | 30 August 2016, 08:46:22 UTC | [FLINK-4526][yarn] remove redundant proxy messages This closes #2437 | 29 August 2016, 16:20:30 UTC |
4cdeb11 | Maximilian Michels | 19 August 2016, 12:57:14 UTC | [FLINK-4486] detached YarnSession: wait until cluster startup is complete This closes #2423 | 29 August 2016, 16:20:00 UTC |
28da995 | Maximilian Michels | 17 August 2016, 13:41:12 UTC | [FLINK-4488][yarn] only automatically shutdown clusters for detached jobs - add check to yarn tests to verify cluster hasn't been shutdown prematurely This closes #2419 | 29 August 2016, 16:18:56 UTC |
3cf494a | Jendrik Poloczek | 29 August 2016, 10:10:28 UTC | [hotfix][docs] fixed timestamp extractor documentation Java example This closes #2433 | 29 August 2016, 14:18:33 UTC |
b3ad028 | Robert Metzger | 29 August 2016, 09:52:48 UTC | [hotfix][quickstarts] Remove Jodatime exclusion | 29 August 2016, 09:52:48 UTC |
81f30c5 | Gordon Tai | 24 August 2016, 08:38:06 UTC | [FLINK-4341] Let idle consumer subtasks emit max value watermarks and fail on resharding This no longer allows the Kinesis consumer to transparently handle resharding. This is a short-term workaround until we have a min-watermark notification service available in the JobManager. This closes #2414 | 29 August 2016, 09:48:38 UTC |
717fc90 | danielblazevski | 13 August 2016, 22:57:18 UTC | [FLINK-3899] [docs] Add examples for incremental window computation. This closes #2368 | 25 August 2016, 12:48:43 UTC |
016c381 | Gyula Fora | 24 August 2016, 10:02:51 UTC | [FLINK-4471] Use user code classloader to deserialize state descriptor in RocksDb backend | 24 August 2016, 16:04:34 UTC |
a0ffa38 | Gyula Fora | 22 August 2016, 05:38:18 UTC | [FLINK-4441] Make RocksDB backend return null on empty state + add test for all backends Closes #2399 | 24 August 2016, 16:02:37 UTC |
b87165d | Maximilian Michels | 24 August 2016, 15:30:36 UTC | [docs] adapt baseURL of documentation | 24 August 2016, 15:30:36 UTC |
70b818b | Maximilian Michels | 22 August 2016, 16:11:45 UTC | [FLINK-4454] always display JobManager address using LeaderRetrievalService This closes #2406 | 24 August 2016, 09:30:35 UTC |
e352ce4 | Ivan Mushketyk | 23 July 2016, 09:39:24 UTC | [FLINK-3866] StringArraySerializer does not treat type as immutable. | 23 August 2016, 14:28:22 UTC |
54cccb8 | Stephan Ewen | 17 August 2016, 09:53:57 UTC | [FLINK-4409] [build] Exclude JSR 305 from Hadoop dependencies | 17 August 2016, 12:59:23 UTC |
cbce10d | zentol | 10 August 2016, 14:23:54 UTC | [FLINK-4302] [metrics] Add documentation to MetricConfig | 17 August 2016, 12:02:30 UTC |
445eaa9 | zentol | 17 August 2016, 11:47:30 UTC | [FLINK-4412] [py] Chaining properly handles broadcast variables | 17 August 2016, 11:53:17 UTC |
aa50e22 | zentol | 17 August 2016, 10:15:37 UTC | [FLINK-4411] [py] Properly propagate chained dual input children | 17 August 2016, 11:53:13 UTC |
c2906c0 | Sunny T | 14 August 2016, 22:50:08 UTC | [FLINK-4309] Fix potential NPE in DelegatingConfiguration This closes #2371. | 16 August 2016, 10:41:39 UTC |
f67cb89 | Maximilian Michels | 10 August 2016, 10:08:23 UTC | [FLINK-4236] fix error handling for jar files with no main method This change shows only entry classes with a valid main method. It ignores classes which contain no main method. This closes #2347 | 13 August 2016, 16:07:08 UTC |
58c810a | Ufuk Celebi | 11 August 2016, 07:55:22 UTC | [docs] Update quickstart version to 1.1.1 | 11 August 2016, 07:55:40 UTC |
68c1a90 | Ufuk Celebi | 11 August 2016, 07:51:45 UTC | [docs] Remove deprecated variables | 11 August 2016, 07:55:40 UTC |
a09f87c | Robert Metzger | 09 August 2016, 11:52:34 UTC | [hotfix][build] Fix release script | 09 August 2016, 12:01:57 UTC |
fc5b58d | Shannon Carey | 08 August 2016, 20:04:23 UTC | [FLINK-4334] [quickstarts] Correctly exclude hadoop1 in quickstart POMs This closes #2341 | 09 August 2016, 10:51:30 UTC |
19de8ec | Stephan Ewen | 08 August 2016, 17:16:23 UTC | [FLINK-4332] [checkpoints] Fix SavepointV1Serializer read() / readFully() | 08 August 2016, 17:29:55 UTC |
b054da4 | Robert Metzger | 08 August 2016, 10:19:03 UTC | [hotfix][kinesis] Fix build instructions | 08 August 2016, 10:23:51 UTC |
e4f62d3 | Maximilian Michels | 01 August 2016, 13:19:15 UTC | [FLINK-4297] [yarn] Decode URL encoded fat jar path This solves problems with spaces and special characters in the automatically determined fat jar path which is returned URL encoded. This closes #2320 | 05 August 2016, 18:03:10 UTC |
5184d0b | Ufuk Celebi | 05 August 2016, 09:49:42 UTC | [FLINK-4314] [tests] Fix test instability in JobManagerHAJobGraphRecoveryITCase The test was relying on the JobManager shutting down before the TaskManager, which is not necessarily the case. If the TaskManager shuts down before the JobManager, the JobGraph could reach the final state FAILED, in which case all HA state is removed. To circumvent this, we add a restart strategy. | 05 August 2016, 09:53:25 UTC |
5363e7a | Greg Hogan | 14 July 2016, 17:40:35 UTC | [FLINK-4219] [scripts] Quote PDSH opts in start-cluster.sh This prevents word splitting if the user configures multiple SSH options. This closes #2253. | 05 August 2016, 08:34:04 UTC |
7768bb2 | Stephan Ewen | 01 August 2016, 14:55:02 UTC | [FLINK-4290] [Cassandra Connector] Skip CassandraConnectorTest on Java 7 builds Cassandra needs Java 8 to run reliably. | 02 August 2016, 18:24:37 UTC |
34eefa3 | Stephan Ewen | 01 August 2016, 12:15:40 UTC | [hotfix] [build] Update LICENSE file with new URL of simplejmx dependency | 02 August 2016, 18:24:37 UTC |
2ea4dd8 | Stephan Ewen | 01 August 2016, 10:17:35 UTC | [hotfix] Remove various '.hidden' files that seem to have been accidentally committed | 02 August 2016, 18:24:37 UTC |
a4d9f57 | philippgrulich | 21 July 2016, 20:31:24 UTC | [FLINK-4251] [Rabbit MQ] Allow users to override queue setup in order to customize queue config This closes #2281 | 02 August 2016, 18:24:36 UTC |
60b1085 | Stephan Ewen | 01 August 2016, 10:03:33 UTC | [FLINK-4292] [batch connectors] Fix setup of HCatalog project by adding Scala SDK dependency | 02 August 2016, 18:24:36 UTC |
a192679 | Stephan Ewen | 01 August 2016, 09:22:27 UTC | [FLINK-4285] [docs] Update the setup quickstart guide with the new SocketWindowWordCount example | 02 August 2016, 18:24:36 UTC |
49e66be | Stephan Ewen | 01 August 2016, 08:01:35 UTC | [hotfix] [docs] Remove obsolete file 'plotPoints.py' | 02 August 2016, 18:24:36 UTC |
e09a454 | Maximilian Michels | 01 August 2016, 16:15:56 UTC | [FLINK-4299] show loss of job manager in Client This prints a message when the leading JobManager changes after first connecting to a JobManager. Further, it prints a message when a connection to a JobManager has been established. This closes #2322. | 02 August 2016, 18:24:36 UTC |
2495184 | Till Rohrmann | 01 August 2016, 16:05:14 UTC | [FLINK-4296] Fixes failure reporting of consumer task scheduling when producer has already finished This PR changes the failure behaviour such that the consumer task is failed instead of the producer task. The latter is problematic, since a finsihed producer task will simply swallow scheduling exception originating from scheduling the consumer task. This closes #2321. | 02 August 2016, 18:24:36 UTC |
383a1cc | Ufuk Celebi | 01 August 2016, 09:13:27 UTC | [FLINK-4289] Unset -x flag for .java files Some source files had the -x flag set: Before this change: ``` $ find . -perm +111 -type f | grep "\.java" ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/places/Attributes.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/places/BoundingBox.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/places/Places.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/Contributors.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/Coordinates.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/CurrentUserRetweet.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/entities/Entities.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/entities/HashTags.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/entities/Media.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/entities/Size.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/entities/Symbol.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/entities/URL.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/entities/UserMention.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/tweet/Tweet.java ./flink-contrib/flink-tweet-inputformat/src/main/java/org/apache/flink/contrib/tweetinputformat/model/User/Users.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/Graph.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/gsa/ApplyFunction.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/gsa/GatherFunction.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/gsa/GatherSumApplyIteration.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/gsa/Neighbor.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/gsa/SumFunction.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/library/GSAConnectedComponents.java ./flink-libraries/flink-gelly/src/main/java/org/apache/flink/graph/library/GSASingleSourceShortestPaths.java ./flink-libraries/flink-gelly-examples/src/main/java/org/apache/flink/graph/examples/GSASingleSourceShortestPaths.java ./flink-libraries/flink-gelly-examples/src/test/java/org/apache/flink/graph/test/GatherSumApplyITCase.java ./flink-runtime/src/main/java/org/apache/flink/runtime/executiongraph/ExecutionGraph.java ``` After this change: ``` $ find . -perm +111 -type f | grep "\.java" ``` | 02 August 2016, 18:24:35 UTC |
67036d8 | Maximilian Michels | 28 July 2016, 12:56:21 UTC | [docs] require at least Ruby 1.9.0 for building The json module is bundled with Ruby from 1.9.0 and upwards. This cures dependency problems with different versions of Ruby. | 02 August 2016, 18:24:35 UTC |
c71a0c7 | Stephan Ewen | 02 August 2016, 16:09:14 UTC | [FLINK-4307] [streaming API] Restore ListState behavior for user-facing ListStates | 02 August 2016, 18:24:35 UTC |
ac7028e | Ufuk Celebi | 01 August 2016, 15:57:18 UTC | Revert "[FLINK-4154] [core] Correction of murmur hash breaks backwards compatibility" This reverts commit 81cf2296683a473db4061dd3bed0aeb249e05058. We had an incorrent implementation of Murmur hash in Flink 1.0. This was fixed in 641a0d4 for Flink 1.1. Then we thought that we need to revert this in order to ensure backwards compatability between Flink 1.0 and 1.1 savepoints (81cf22). Turns out, savepoint backwards compatability is broken for other reasons, too. Therefore, we revert 81cf22 here, ending up with a correct implementation of Murmur hash again. | 02 August 2016, 18:24:35 UTC |
12bf7c1 | kl0u | 18 July 2016, 09:37:06 UTC | [FLINK-4207] WindowOperator becomes very slow with allowed lateness | 26 July 2016, 19:12:05 UTC |
884d3e2 | Maximilian Michels | 26 July 2016, 16:35:33 UTC | [FLINK-4258] fix potential NPE in SavepointCoordinator | 26 July 2016, 16:37:58 UTC |
9bcbcf4 | Aljoscha Krettek | 22 July 2016, 13:25:01 UTC | [FLINK-4246] Allow Specifying Multiple Metrics Reporters This also updates documentation and tests. Reporters can now be specified like this: metrics.reporters: foo,bar metrics.reporter.foo.class: JMXReporter.class metrics.reporter.foo.port: 10 metrics.reporter.bar.class: GangliaReporter.class metrics.reporter.bar.port: 11 metrics.reporter.bar.something: 42 | 26 July 2016, 15:05:21 UTC |
cd232e6 | Maximilian Michels | 21 July 2016, 15:06:06 UTC | [FLINK-4261][tools] retry in case of failed snapshot deployment Unfortunately, we can't deploy snapshots atomically using the Nexus repository. The staged process which leads to an atomic deployment is only designed to work for releases. Best we can do is to retry deploying artifacts in case of failures. - introduce retry in case of failure of snapshot deployment - simplify deployment script This closes #2296 | 26 July 2016, 14:46:47 UTC |
2648bc1 | Till Rohrmann | 15 July 2016, 08:51:59 UTC | [FLINK-4152] Allow re-registration of TMs at resource manager - Add YarnFlinkResourceManager test to reaccept task manager registrations from a re-elected job manager - Remove unnecessary sync logic between JobManager and ResourceManager - Avoid duplicate reigstration attempts in case of a refused registration - Add test case to check that not an excessive amount of RegisterTaskManager messages are sent - Remove containersLaunched from YarnFlinkResourceManager and instead not clearing registeredWorkers when JobManager loses leadership - Let YarnFlinkResourceManagerTest extend TestLogger - Harden YarnFlinkResourceManager.getContainersFromPreviousAttempts - Add FatalErrorOccurred message handler to FlinkResourceManager; Increase timeout for YarnFlinkResourceManagerTest; Add additional constructor to TestingYarnFlinkResourceManager for tests - Rename registeredWorkers field into startedWorkers Additionally, the RegisterResource message is renamed into NotifyResourceStarted which tells the RM that a resource has been started. This reflects the current semantics of the startedWorkers map in the resource manager. - Fix concurrency issues in TestingLeaderRetrievalService This closes #2257 | 26 July 2016, 14:39:22 UTC |
c6715b7 | Maximilian Michels | 22 July 2016, 15:06:06 UTC | [travis] remove outdated comment regarding snapshots | 26 July 2016, 11:46:39 UTC |
e3fec1f | zentol | 22 July 2016, 10:33:16 UTC | [FLINK-4192] [metrics] Move metrics classes out of 'flink-core' - moved user-facing API to 'flink-metrics/flink-metrics-core' - moved JMXReporter to 'flink-metrics/flink-metrics-jmx' - moved remaining metric classes to 'flink-runtime' This closes #2226 | 26 July 2016, 10:30:15 UTC |
e4fe89d | zentol | 22 July 2016, 13:20:08 UTC | [FLINK-4210][metrics] Move close()/isClosed() out of MetricGroup This closes #2286 | 26 July 2016, 09:50:01 UTC |
0b4c04d | Aljoscha Krettek | 21 July 2016, 13:10:54 UTC | [FLINK-4239] Set Default Allowed Lateness to Zero and Make Triggers Non-Purging This closes #2278. | 26 July 2016, 09:32:42 UTC |
f0ac261 | Ufuk Celebi | 30 June 2016, 14:16:41 UTC | [FLINK-4067] [runtime] Add savepoint headers Savepoints were previously persisted without any meta data using default Java serialization of `CompletedCheckpoint`. This commit introduces a savepoint interface with version-specific serializers and stores savepoints with meta data. Savepoints expose a version number and a Collection<TaskState> for savepoint restore. Currently, there is only one savepoint version: SavepointV0 (Flink 1.1): This is the current savepoint version, which holds a reference to the Checkpoint task state collection, but is serialized with a custom serializater not relying on default Java serialization. Therefore, it should not happen again that we need to stick to certain classes in future Flink versions. The savepoints are stored in `FsSavepointStore` with the following format: MagicNumber SavepointVersion Savepoint - MagicNumber => int - SavepointVersion => int (returned by Savepoint#getVersion()) - Savepoint => bytes (serialized via version-specific SavepointSerializer) The header is minimal (magic number, version). All savepoint-specific meta data can be moved to the savepoint itself. This is also were we would have to add new meta data in future versions, allowing us to differentiate between different savepoint versions when we change the serialization stack. All savepoint related classes have been moved from checkpoint to a new sub package `checkpoint.savepoint`. This closes #2194. | 26 July 2016, 09:30:43 UTC |
110bba3 | twalthr | 26 July 2016, 08:44:24 UTC | [FLINK-4103] [table] Add CsvTableSource docs and Java accessibility | 26 July 2016, 09:03:29 UTC |
7e309ee | smarthi | 25 June 2016, 03:55:14 UTC | [FLINK-4103] [table] Modify CsvTableSource to implement StreamTableSource This closes #2162. | 26 July 2016, 08:47:57 UTC |
0360fb9 | Greg Hogan | 25 July 2016, 15:24:10 UTC | [hotfix] [gelly] Reduce maximum number of generator blocks The default maximum akka transfer size is 10 MB. This commit reduces the number of generator blocks from 2^20 to 2^15 which removes the limit on graph size. The original limit of one millions blocks was intended to future-proof scalability. This is a temporary fix as graph generation will be reworked in FLINK-3997. | 25 July 2016, 17:58:21 UTC |
177168b | erli ding | 13 July 2016, 03:29:34 UTC | [FLINK-2125][streaming] Delimiter change from char to string This closes #2233 | 25 July 2016, 15:37:44 UTC |
588830a | Jark Wu | 21 July 2016, 15:25:44 UTC | [FLINK-4244] [docs] Field names for union operator do not have to be equal This closes #2280. | 25 July 2016, 13:47:07 UTC |
b6f9308 | twalthr | 25 July 2016, 08:47:53 UTC | [FLINK-3891] [table] Add a class containing all supported Table API types This closes #2292. | 25 July 2016, 13:31:21 UTC |
2f9a28a | twalthr | 22 July 2016, 09:26:43 UTC | [FLINK-3901] [table] Convert Java implementation to Scala and fix bugs This closes #2283. | 25 July 2016, 13:23:38 UTC |
c5d1d12 | Flavio Pompermaier | 13 May 2016, 16:19:51 UTC | [FLINK-3901] [table] Create a RowCsvInputFormat to use as default CSV IF in Table API | 25 July 2016, 13:23:38 UTC |
3213016 | Ufuk Celebi | 14 July 2016, 14:29:49 UTC | [FLINK-4150] [runtime] Don't clean up BlobStore on BlobServer shut down The `BlobServer` acts as a local cache for uploaded BLOBs. The life-cycle of each BLOB is bound to the life-cycle of the `BlobServer`. If the BlobServer shuts down (on JobManager shut down), all local files will be removed. With HA, BLOBs are persisted to another file system (e.g. HDFS) via the `BlobStore` in order to have BLOBs available after a JobManager failure (or shut down). These BLOBs are only allowed to be removed when the job that requires them enters a globally terminal state (`FINISHED`, `CANCELLED`, `FAILED`). This commit removes the `BlobStore` clean up call from the `BlobServer` shutdown. The `BlobStore` files will only be cleaned up via the `BlobLibraryCacheManager`'s' clean up task (periodically or on BlobLibraryCacheManager shutdown). This means that there is a chance that BLOBs will linger around after the job has terminated, if the job manager fails before the clean up. This closes #2256. | 25 July 2016, 13:20:26 UTC |
2818ee0 | zentol | 22 July 2016, 15:50:10 UTC | [hotfix] Prevent CheckpointCommitter from failing job Prevents the CheckpointCommitter from failing a job, if either commitCheckpoint() or isCheckpointCommitter() failed. Instead, we will try again on the next notify(). This closes #2287 | 25 July 2016, 12:52:55 UTC |
b71ac35 | Greg Hogan | 14 July 2016, 14:02:23 UTC | [FLINK-4217] [gelly] Gelly drivers should read CSV values as strings The user must now specify the ID type as "integer" or "string" when reading a graph from a CSV file. This closes #2250 | 22 July 2016, 18:07:19 UTC |
e2ef74e | Greg Hogan | 14 July 2016, 13:39:02 UTC | [FLINK-4213] [gelly] Provide CombineHint in Gelly algorithms This closes #2248 | 22 July 2016, 18:07:19 UTC |
54f02ec | Ufuk Celebi | 19 July 2016, 14:30:23 UTC | [FLINK-4201] [runtime] Forward suspend to checkpoint coordinator Suspended jobs were leading to shutdown of the checkpoint coordinator and hence removal of checkpoint state. For standalone recovery mode this is OK as no state can be recovered anyways (unchanged in this PR). For HA though this lead to removal of checkpoint state, which we actually want to keep for recovery. We have the following behaviour now: -----------+------------+------------------- | Standalone | High Availability -----------+------------+------------------- SUSPENDED | Discard | Keep -----------+------------+------------------- FINISHED/ | Discard | Discard FAILED/ | | CANCELED | | -----------+------------+------------------- This closes #2276. | 22 July 2016, 12:26:40 UTC |
b3fa459 | Till Rohrmann | 22 July 2016, 10:41:51 UTC | [FLINK-4202] [metrics] Update restarting time metric documentation This closes #2284. | 22 July 2016, 12:18:52 UTC |
73836f7 | Aljoscha Krettek | 21 July 2016, 14:44:06 UTC | [FLINK-4229] [metrics] Only start JMX server when port is specified This closes #2279 | 21 July 2016, 15:02:51 UTC |
5dd85e2 | Till Rohrmann | 20 July 2016, 15:08:18 UTC | [FLINK-4202] Add restarting time JM metric This PR adds a JM metric which shows the time it took to restart a job. The time is measured between entering the JobStatus.RESTARTING and reaching the JobStatus.RUNNING state. During this time, the restarting time is continuously updated. The metric only shows the time for the last restart attempt. The metric is published in the job metric group under the name of "restartingTime". This closes #2271. | 21 July 2016, 14:16:31 UTC |
d3bc556 | Aljoscha Krettek | 20 July 2016, 16:10:11 UTC | [FLINK-4238] Only allow/require query for Tuple Stream in CassandraSink | 21 July 2016, 13:21:13 UTC |
f81cda3 | Ufuk Celebi | 20 July 2016, 14:15:22 UTC | [FLINK-4237] [runtime] Cancel savepoints on declined snapshots | 21 July 2016, 11:50:12 UTC |
ccd4fd9 | Stefan Richter | 20 July 2016, 10:20:23 UTC | [FLINK-4230] [DataStreamAPI] PR feedback on Session Windowing ITCase | 21 July 2016, 08:11:30 UTC |
78af1e9 | Stefan Richter | 17 July 2016, 16:31:13 UTC | [FLINK-4230] [DataStreamAPI] Add Session Windowing ITCase | 21 July 2016, 08:11:30 UTC |
19dae21 | Aljoscha Krettek | 18 July 2016, 15:46:21 UTC | [FLINK-4229] Do not start any Metrics Reporter by default | 20 July 2016, 16:02:48 UTC |
25b6f22 | kl0u | 18 July 2016, 16:18:43 UTC | [FLINK-3956] Make FileInputFormats independent from Configuration Parameters of some input formats that was only possible to be set through the Configuration object now have setter methods that allow the user to do so. Values set by the setters cannot be reset by the configuration object. | 20 July 2016, 15:59:35 UTC |
6ab96a7 | zentol | 13 July 2016, 11:51:05 UTC | [FLINK-3729] [table] Several SQL tests fail on Windows OS This closes #2238. | 20 July 2016, 14:25:53 UTC |
130511f | Jark Wu | 17 June 2016, 03:39:21 UTC | [FLINK-4070] [table] Support literals on left side of binary expressions This closes #2120. | 20 July 2016, 13:01:47 UTC |
2391683 | kl0u | 15 July 2016, 16:20:04 UTC | [hotfix] Fixes broken TopSpeedWindowing scala example Integrated PR comments This closes #2259. | 20 July 2016, 12:24:06 UTC |
082d87e | Stefan Richter | 14 July 2016, 16:21:40 UTC | [FLINK-4166] [Distributed Coordination] zookeeper namespaces (cli parameter -z) This closes #2249 | 19 July 2016, 16:00:21 UTC |
17589d4 | Maximilian Michels | 19 July 2016, 07:51:23 UTC | [FLINK-4199] fix misleading CLI messages during job submission - change CLI message upon cluster retrieval - save JobExecutionResult for interactive executions - only print Collection size in accumulator results - remove unused helper method This closes #2264 | 19 July 2016, 16:00:04 UTC |
e85f787 | twalthr | 08 July 2016, 14:50:17 UTC | [FLINK-4183] [table] Move checking for StreamTableEnvironment into validation layer This closes #2221. | 19 July 2016, 14:27:35 UTC |