| 0c37622 | Matei Zaharia | 07 February 2013, 06:35:26 UTC | Update version number to 0.6.2 | 07 February 2013, 06:35:26 UTC |
| deb9264 | Matei Zaharia | 07 February 2013, 05:56:00 UTC | Update Windows scripts to launch daemons with less RAM and fix a few other issues | 07 February 2013, 05:56:00 UTC |
| 98c5ad2 | Stephen Haberman | 08 January 2013, 20:44:33 UTC | Retrieve jars to a flat directory so * can be used for the classpath. Conflicts: run | 07 February 2013, 05:43:30 UTC |
| f842372 | Matei Zaharia | 07 February 2013, 02:13:32 UTC | Update tutorial to list Akka repository instead of Typesafe | 07 February 2013, 02:13:32 UTC |
| c0d2ea1 | Matei Zaharia | 06 February 2013, 22:34:46 UTC | Use a separate memory setting for standalone cluster daemons | 06 February 2013, 22:34:46 UTC |
| caace6e | Matei Zaharia | 06 February 2013, 22:14:14 UTC | Tweak web UI so that people don't get confused about master URL format | 06 February 2013, 22:14:14 UTC |
| f886b42 | Matei Zaharia | 06 February 2013, 21:30:35 UTC | Some bug and formatting fixes to FT | 06 February 2013, 21:30:35 UTC |
| 6a5c389 | Matei Zaharia | 06 February 2013, 20:38:32 UTC | Change dependency to Akka repo to deal with a Maven dependency problem (in resolving commons-httpclient 3.0.1) | 06 February 2013, 20:38:32 UTC |
| e395aa2 | root | 06 February 2013, 19:27:20 UTC | Increase worker timeouts | 06 February 2013, 19:27:20 UTC |
| 171165b | root | 06 February 2013, 19:25:55 UTC | Properly deal with removed hosts in BlockManagerMaster | 06 February 2013, 19:25:55 UTC |
| 4b53f14 | root | 06 February 2013, 09:11:17 UTC | Detect hard crashes of workers using a heartbeat mechanism. Also fixes some issues in the rest of the code with detecting workers this way. | 06 February 2013, 09:11:17 UTC |
| 12f6107 | Matei Zaharia | 04 February 2013, 23:03:31 UTC | Fix compilation due to previous cherry-pick | 04 February 2013, 23:03:31 UTC |
| 93298bf | Charles Reiss | 30 January 2013, 01:09:53 UTC | Remember ConnectionManagerId used to initiate SendingConnections. This prevents ConnectionManager from getting confused if a machine has multiple host names and the one getHostName() finds happens not to be the one that was passed from, e.g., the BlockManagerMaster. | 04 February 2013, 20:33:36 UTC |
| d079487 | Charles Reiss | 11 December 2012, 05:40:09 UTC | Call slaveLost on executor death for standalone clusters. | 04 February 2013, 20:25:19 UTC |
| 2f36810 | Charles Reiss | 23 January 2013, 20:59:51 UTC | Remove more dead code from test. | 04 February 2013, 20:08:44 UTC |
| bb34ad2 | Charles Reiss | 23 January 2013, 20:48:45 UTC | Clarify TODO. | 04 February 2013, 20:08:38 UTC |
| ed4bae5 | Charles Reiss | 23 January 2013, 20:40:38 UTC | Remove dead code in test. | 04 February 2013, 20:08:35 UTC |
| 81cd647 | Charles Reiss | 23 January 2013, 09:37:51 UTC | Add tests of various node failure scenarios. | 04 February 2013, 20:08:30 UTC |
| 64c8fc5 | Charles Reiss | 23 January 2013, 09:35:14 UTC | Extra debugging from hostLost() | 04 February 2013, 20:08:25 UTC |
| 59b5647 | Charles Reiss | 23 January 2013, 09:34:44 UTC | Force generation increment after shuffle map stage | 04 February 2013, 20:08:14 UTC |
| 5e6c9a5 | Charles Reiss | 22 January 2013, 08:23:31 UTC | Use generation numbers for fetch failure tracking Conflicts: core/src/main/scala/spark/scheduler/DAGScheduler.scala | 04 February 2013, 20:07:57 UTC |
| 1461774 | Matei Zaharia | 01 February 2013, 23:35:12 UTC | Merge pull request #440 from pwendell/branch-0.6 Typo in Maven path (0.6 branch) | 01 February 2013, 23:35:12 UTC |
| c3aa083 | Patrick Wendell | 01 February 2013, 23:27:52 UTC | Typo in Maven path (0.6 branch) | 01 February 2013, 23:27:52 UTC |
| 16a9682 | Josh Rosen | 26 January 2013, 23:57:01 UTC | Fix JavaRDDLike.flatMap(PairFlatMapFunction) (SPARK-668). This workaround is easier than rewriting JavaRDDLike in Java. | 27 January 2013, 00:32:02 UTC |
| c2490a2 | Matei Zaharia | 23 January 2013, 00:55:55 UTC | Merge pull request #393 from squito/kryo-class-loader be sure to set class loader of kryo instances | 23 January 2013, 00:55:55 UTC |
| e6cc9f0 | Stephen Haberman | 22 January 2013, 21:30:42 UTC | Handle slaveLost before slaveIdToHost knows about it. Conflicts: core/src/main/scala/spark/scheduler/cluster/ClusterScheduler.scala | 23 January 2013, 00:35:03 UTC |
| e0c2891 | Leemoonsoo | 22 January 2013, 14:08:34 UTC | Fix for hanging spark.HttpFileServer with kind of virtual network | 22 January 2013, 23:50:15 UTC |
| 3e8d903 | Imran Rashid | 22 January 2013, 00:21:14 UTC | be sure to set class loader of kryo instances | 22 January 2013, 00:21:23 UTC |
| 149d477 | Matei Zaharia | 21 January 2013, 21:10:02 UTC | Add more Akka settings to docs | 21 January 2013, 21:10:02 UTC |
| 55aa052 | Matei Zaharia | 21 January 2013, 21:02:40 UTC | Clarify the documentation on env variables for standalone mode | 21 January 2013, 21:02:40 UTC |
| 44f4be2 | Matei Zaharia | 21 January 2013, 20:37:41 UTC | Merge branch 'branch-0.6' of github.com:mesos/spark into branch-0.6 | 21 January 2013, 20:37:41 UTC |
| b804e8e | Matei Zaharia | 21 January 2013, 19:59:21 UTC | Only bind to IPv4 addresses when trying to auto-detect external IP | 21 January 2013, 20:00:28 UTC |
| 2a90ea2 | Matei Zaharia | 21 January 2013, 19:20:42 UTC | Merge pull request #384 from squito/executor_spark_home executor can use a different sparkHome from Worker | 21 January 2013, 19:20:42 UTC |
| ec4c550 | Imran Rashid | 21 January 2013, 18:52:24 UTC | Merge branch 'branch-0.6' into executor_spark_home | 21 January 2013, 18:52:24 UTC |
| 255b48f | Imran Rashid | 21 January 2013, 18:52:17 UTC | more File -> String changes | 21 January 2013, 18:52:17 UTC |
| bf7143b | Matei Zaharia | 21 January 2013, 08:23:53 UTC | Log remote lifecycle events in Akka for easier debugging | 21 January 2013, 08:24:44 UTC |
| 7de03d8 | Imran Rashid | 21 January 2013, 05:57:44 UTC | remove unused imports | 21 January 2013, 05:57:44 UTC |
| 150b86d | Imran Rashid | 21 January 2013, 05:55:50 UTC | send sparkHome as String instead of File over network | 21 January 2013, 05:55:50 UTC |
| 4c12344 | Imran Rashid | 19 January 2013, 21:26:15 UTC | oops, fix stupid compile error | 19 January 2013, 21:26:15 UTC |
| 1c6aecf | Imran Rashid | 18 January 2013, 21:28:50 UTC | get sparkHome the correct way | 18 January 2013, 21:28:50 UTC |
| f54e4e2 | Imran Rashid | 18 January 2013, 21:24:37 UTC | executor can use a different sparkHome from Worker | 18 January 2013, 21:24:37 UTC |
| 52a80d1 | Matei Zaharia | 18 January 2013, 01:40:55 UTC | Fix code that assumed spark.local.dir is only a single directory Conflicts: core/src/main/scala/spark/Utils.scala | 18 January 2013, 01:45:52 UTC |
| 852c0a2 | Fernand Pajot | 18 January 2013, 00:56:11 UTC | changed HttpBroadcast server cache to be in spark.local.dir instead of java.io.tmpdir | 18 January 2013, 01:02:07 UTC |
| fede00d | Imran Rashid | 16 January 2013, 20:21:37 UTC | remove unused thread pool | 17 January 2013, 18:36:08 UTC |
| f103e9d | Josh Rosen | 30 December 2012, 20:43:06 UTC | Raise exception when hashing Java arrays (SPARK-597) | 17 January 2013, 18:35:33 UTC |
| 4cf77ad | Charles Reiss | 15 January 2013, 20:05:54 UTC | Remove broken attempt to test fetching case. | 16 January 2013, 00:05:13 UTC |
| 3f70ae6 | Charles Reiss | 15 January 2013, 01:04:44 UTC | Fix accidental spark.master.host reuse | 16 January 2013, 00:04:56 UTC |
| 4b071ce | Charles Reiss | 14 January 2013, 23:24:08 UTC | Additional tests for MapOutputTracker. | 16 January 2013, 00:04:51 UTC |
| 18dfd37 | Charles Reiss | 03 January 2013, 22:09:56 UTC | Throw FetchFailedException for cached missing locs | 16 January 2013, 00:04:44 UTC |
| f7c6cbd | Andrew Psaltis | 15 January 2013, 19:45:42 UTC | Changed teh scala version to 2.9.2, so that the classes can be found when the classpath is expanded. | 15 January 2013, 22:21:44 UTC |
| afd7564 | Josh Rosen | 14 December 2012, 05:53:30 UTC | Add try-finally to handle MapOutputTracker timeouts. Conflicts: core/src/main/scala/spark/MapOutputTracker.scala | 15 January 2013, 18:31:06 UTC |
| 76c708a | Matei Zaharia | 14 January 2013, 04:05:20 UTC | Import small optimization to CoGroupedRDD from master (by Reynold Xin) | 14 January 2013, 04:05:20 UTC |
| e7b2e16 | Matei Zaharia | 14 January 2013, 03:34:07 UTC | Make filter preserve partitioner info, since it can Conflicts: core/src/main/scala/spark/rdd/FilteredRDD.scala | 14 January 2013, 03:39:53 UTC |
| 72a73d7 | shane-huang | 10 January 2013, 12:09:34 UTC | Modified Patch according to comments | 11 January 2013, 01:55:24 UTC |
| 94847e4 | shane-huang | 08 January 2013, 14:40:58 UTC | Fix an issue in ConnectionManager where sendingMessage may create too many unnecessary SendingConnections. | 11 January 2013, 01:55:16 UTC |
| fff3266 | Josh Rosen | 14 December 2012, 01:19:00 UTC | Use spark-env.sh to configure standalone master. See SPARK-638. Also fixed a typo in the standalone mode documentation. | 14 December 2012, 02:13:27 UTC |
| 6741bc2 | Matei Zaharia | 14 December 2012, 02:12:45 UTC | Import EC2 script updates from master branch | 14 December 2012, 02:12:45 UTC |
| 46555cc | Peter Sankauskas | 11 December 2012, 01:44:09 UTC | SPARK-626: Remove rules before removing security groups, with a pause in between so wait for AWS eventual consistency to catch up. Conflicts: ec2/spark_ec2.py | 14 December 2012, 02:09:53 UTC |
| 5867a24 | Matei Zaharia | 10 December 2012, 23:12:06 UTC | Make sure the SSH key we copy to EC2 has permissions 600. SPARK-539 #resolve | 10 December 2012, 23:19:38 UTC |
| 5dd9498 | Matei Zaharia | 08 December 2012, 08:33:11 UTC | Search for a non-loopback address in Utils.getLocalIpAddress | 08 December 2012, 09:07:02 UTC |
| d7cf579 | Josh Rosen | 02 December 2012, 02:18:15 UTC | Use external addresses in standalone WebUI on EC2. | 04 December 2012, 23:36:09 UTC |
| 4d54aaa | Patrick Wendell | 28 November 2012, 07:30:10 UTC | Adding multi-jar constructor in quickstart | 29 November 2012, 07:01:22 UTC |
| 4868f97 | Matei Zaharia | 28 November 2012, 05:00:04 UTC | Use hostname instead of IP in deploy scripts to let Akka connect properly | 28 November 2012, 05:00:40 UTC |
| 314a86a | Josh Rosen | 12 November 2012, 01:45:25 UTC | Fix NullPointerException caused by unregistered map outputs. | 28 November 2012, 04:55:43 UTC |
| cfd9d40 | Charles Reiss | 27 November 2012, 21:55:56 UTC | Fix NullPointerException when unregistering a map output twice. | 28 November 2012, 04:55:33 UTC |
| 87d5327 | Charles Reiss | 27 November 2012, 20:50:40 UTC | Tests for MapOutputTracker. | 28 November 2012, 04:55:27 UTC |
| 005f1ff | Matei Zaharia | 23 November 2012, 00:25:29 UTC | Change version to 0.6.2-SNAPSHOT | 23 November 2012, 00:25:35 UTC |
| edb91a3 | mbautin | 19 November 2012, 19:52:10 UTC | Addressing Matei's comment: SPARK_LOCAL_IP environment variable | 19 November 2012, 21:44:51 UTC |
| eeb093b | mbautin | 15 November 2012, 21:44:13 UTC | SPARK-624: make the default local IP customizable | 19 November 2012, 21:44:42 UTC |
| 5c0835f | Matei Zaharia | 17 November 2012, 04:48:35 UTC | Doc fix | 17 November 2012, 04:48:42 UTC |
| 7c1be97 | Matei Zaharia | 15 November 2012, 20:55:34 UTC | Change version to 0.6.1 | 15 November 2012, 20:55:34 UTC |
| 8892603 | Matei Zaharia | 15 November 2012, 08:10:52 UTC | Use DNS names instead of IP addresses in standalone mode, to allow matching with data locality hints from storage systems. | 15 November 2012, 08:11:40 UTC |
| 9d8e17d | Matei Zaharia | 15 November 2012, 06:37:05 UTC | Fix Java/Scala home having spaces on Windows | 15 November 2012, 08:11:31 UTC |
| 9b80225 | Matei Zaharia | 12 November 2012, 05:06:57 UTC | Detect correctly when one has disconnected from a standalone cluster. SPARK-617 #resolve | 12 November 2012, 05:22:17 UTC |
| dc84ce7 | root | 11 November 2012, 07:05:22 UTC | Fix K-means example a little | 11 November 2012, 07:05:22 UTC |
| 9d5740f | Tathagata Das | 09 November 2012, 23:46:15 UTC | Incorporated Matei's suggestions. Tested with 5 producer(consumer) threads each doing 50k puts (gets), took 15 minutes to run, no errors or deadlocks. | 10 November 2012, 04:28:04 UTC |
| e870ca5 | Tathagata Das | 09 November 2012, 22:09:37 UTC | Fixed deadlock in BlockManager. 1. Changed the lock structure of BlockManager by replacing the 337 coarse-grained locks to use BlockInfo objects as per-block fine-grained locks. 2. Changed the MemoryStore lock structure by making the block putting threads lock on a different object (not the memory store) thus making sure putting threads minimally blocks to the getting treads. 3. Added spark.storage.ThreadingTest to stress test the BlockManager using 5 block producer and 5 block consumer threads. | 10 November 2012, 04:27:56 UTC |
| bb2b9ff | Matei Zaharia | 09 November 2012, 07:13:12 UTC | Added an option to spread out jobs in the standalone mode. | 09 November 2012, 07:13:50 UTC |
| b3b52c9 | Matei Zaharia | 08 November 2012, 17:53:40 UTC | Fix for connections not being reused (from Josh Rosen) | 08 November 2012, 17:54:17 UTC |
| a245408 | Matei Zaharia | 08 November 2012, 08:10:13 UTC | Merge pull request #300 from enachb/mesos_slavelost fix bug in getting slave id out of mesos | 08 November 2012, 08:10:13 UTC |
| 4fe0d80 | Imran Rashid | 07 November 2012, 23:35:51 UTC | fix bug in getting slave id out of mesos | 08 November 2012, 05:45:54 UTC |
| 5acd753 | Matei Zaharia | 21 October 2012, 06:33:37 UTC | Various fixes to standalone mode and web UI: - Don't report a job as finishing multiple times - Don't show state of workers as LOADING when they're running - Show start and finish times in web UI - Sort web UI tables by ID and time by default | 08 November 2012, 00:51:16 UTC |
| 171e97a | Josh Rosen | 31 October 2012, 06:32:38 UTC | Cancel spot instance requests when exiting spark-ec2. | 07 November 2012, 23:14:32 UTC |
| d20142b | Shivaram Venkataraman | 01 November 2012, 17:46:38 UTC | Remove unnecessary hash-map put in MemoryStore | 07 November 2012, 23:13:59 UTC |
| 43465e9 | Josh Rosen | 04 November 2012, 00:02:47 UTC | Fix check for existing instances during EC2 launch. | 07 November 2012, 23:08:40 UTC |
| cf0bf73 | Josh Rosen | 18 October 2012, 17:01:38 UTC | Allow EC2 script to stop/destroy cluster after master/slave failures. | 07 November 2012, 23:06:00 UTC |
| ef683d4 | Josh Rosen | 23 October 2012, 20:49:52 UTC | Fix minor typos in quick start guide. | 07 November 2012, 23:03:50 UTC |
| 222355e | Thomas Dudziak | 22 October 2012, 20:10:47 UTC | Tweaked run file to live more happily with typesafe's debian package | 07 November 2012, 23:02:46 UTC |
| d2b2fc2 | Matei Zaharia | 06 November 2012, 23:57:38 UTC | Made Akka timeout and message frame size configurable, and upped the defaults | 06 November 2012, 23:57:38 UTC |
| d338742 | root | 26 October 2012, 07:31:08 UTC | Don't throw an error in the block manager when a block is cached on the master due to a locally computed operation | 26 October 2012, 07:31:08 UTC |
| ce143d6 | Matei Zaharia | 25 October 2012, 04:52:13 UTC | Strip leading mesos:// in URLs passed to Mesos | 25 October 2012, 04:52:40 UTC |
| 902a608 | Matei Zaharia | 22 October 2012, 18:43:57 UTC | Update version to 0.6.1-SNAPSHOT to show this is in development | 22 October 2012, 18:43:57 UTC |
| 1be335e | Matei Zaharia | 21 October 2012, 07:05:02 UTC | Merge branch 'master' into dev | 21 October 2012, 07:05:02 UTC |
| 15e95be | Matei Zaharia | 21 October 2012, 06:35:01 UTC | Merge pull request #285 from tomdz/cdh4-dev Support for Hadoop 2 distributions such as cdh4 | 21 October 2012, 06:35:01 UTC |
| 6999724 | Matei Zaharia | 21 October 2012, 06:33:37 UTC | Fix a path in the web UI | 21 October 2012, 06:33:37 UTC |
| 45430f2 | Patrick Wendell | 20 October 2012, 06:18:44 UTC | Merge pull request #290 from pwendell/dev Two trivial commits to test JIRA integration | 20 October 2012, 06:18:44 UTC |
| cd09365 | Patrick Wendell | 20 October 2012, 06:14:37 UTC | SPARK-581 #resolve Removing whitespace to test JIRA | 20 October 2012, 06:17:44 UTC |
| d50028b | Patrick Wendell | 20 October 2012, 06:14:25 UTC | Adding whitespace to test JIRA integration | 20 October 2012, 06:17:44 UTC |
| bff5cef | Matei Zaharia | 19 October 2012, 08:12:05 UTC | Merge pull request #287 from rxin/startslave Use SPARK_MASTER_IP if it is set in start-slaves.sh. | 19 October 2012, 08:12:05 UTC |
| f67bcbe | Reynold Xin | 19 October 2012, 08:08:23 UTC | Use SPARK_MASTER_IP if it is set in start-slaves.sh. | 19 October 2012, 08:08:23 UTC |
