Commit message (Collapse) | Author | Age | Files | Lines | |
---|---|---|---|---|---|
* | Rename top-level 'pyspark' directory to 'python' | Josh Rosen | 2013-01-01 | 1 | -1/+1 |
| | |||||
* | Minor documentation and style fixes for PySpark. | Josh Rosen | 2013-01-01 | 2 | -15/+32 |
| | |||||
* | Update PySpark for compatibility with TaskContext. | Josh Rosen | 2012-12-29 | 1 | -8/+5 |
| | |||||
* | Merge remote-tracking branch 'origin/master' into python-api | Josh Rosen | 2012-12-29 | 97 | -1531/+2892 |
|\ | | | | | | | | | Conflicts: docs/quick-start.md | ||||
| * | Change Utils.fetchFile() warning to SparkException. | Josh Rosen | 2012-12-28 | 2 | -7/+10 |
| | | |||||
| * | Add addFile() and addJar() to JavaSparkContext. | Josh Rosen | 2012-12-28 | 1 | -0/+34 |
| | | |||||
| * | Add synchronization to LocalScheduler.updateDependencies(). | Josh Rosen | 2012-12-28 | 1 | -16/+18 |
| | | |||||
| * | Skip deletion of files in clearFiles(). | Josh Rosen | 2012-12-28 | 2 | -20/+46 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This fixes an issue where Spark could delete original files in the current working directory that were added to the job using addFile(). There was also the potential for addFile() to overwrite local files, which is addressed by changing Utils.fetchFile() to log a warning instead of overwriting a file with new contents. This is a short-term fix; a better long-term solution would be to remove the dependence on storing files in the current working directory, since we can't change the cwd from Java. | ||||
| * | fall back to filter-map-collect when calling lookup() on an RDD without a ↵ | Mark Hamstra | 2012-12-24 | 2 | -1/+12 |
| | | | | | | | | partitioner | ||||
| * | Allow distinct() to be called without parentheses when using the default ↵ | Mark Hamstra | 2012-12-24 | 2 | -5/+11 |
| | | | | | | | | number of splits. | ||||
| * | Remove the call to close input stream in Kryo serializer. | Reynold Xin | 2012-12-21 | 1 | -1/+1 |
| | | |||||
| * | Merge pull request #341 from rxin/4a3fb06ac2d11125feb08acbbd4df76d1e91b677 | Matei Zaharia | 2012-12-21 | 1 | -137/+68 |
| |\ | | | | | | | Kryo2 update against Spark master | ||||
| | * | Updated Kryo to 2.20. | Reynold Xin | 2012-10-16 | 1 | -137/+68 |
| | | | |||||
| * | | Merge pull request #327 from rxin/spark-633 | Matei Zaharia | 2012-12-20 | 16 | -769/+1096 |
| |\ \ | | | | | | | | | Added the ability in block manager to remove blocks. | ||||
| | * | | Let the slave notify the master block removal. | Reynold Xin | 2012-12-20 | 4 | -92/+83 |
| | | | | |||||
| | * | | Moved BlockManager's IdGenerator into BlockManager object. Removed some | Reynold Xin | 2012-12-19 | 5 | -36/+22 |
| | | | | | | | | | | | | | | | | excessive debug messages. | ||||
| | * | | Merge branch 'spark-633' of github.com:rxin/spark into spark-633 | Reynold Xin | 2012-12-14 | 4 | -7/+23 |
| | |\ \ | |||||
| | | * | | Return success or failure in BlockStore.remove(). | Josh Rosen | 2012-12-13 | 4 | -7/+23 |
| | | | | | |||||
| | * | | | Fixed conflicts from merging Charles' and TD's block manager changes. | Reynold Xin | 2012-12-14 | 3 | -173/+158 |
| | | | | | |||||
| | * | | | Merge branch 'master' of github.com:mesos/spark into spark-633 | Reynold Xin | 2012-12-13 | 1 | -7/+9 |
| | |\ \ \ | |||||
| | * | | | | Merged TD's block manager refactoring. | Reynold Xin | 2012-12-13 | 9 | -647/+805 |
| | | | | | | |||||
| | * | | | | Merge branch 'master' of github.com:mesos/spark into spark-633 | Reynold Xin | 2012-12-13 | 8 | -15/+118 |
| | |\ \ \ \ | |||||
| | * \ \ \ \ | Merge branch 'master' of github.com:mesos/spark into spark-633 | Reynold Xin | 2012-12-13 | 26 | -208/+207 |
| | |\ \ \ \ \ | | | |_|_|/ / | | |/| | | | | |||||
| | * | | | | | Added the ability in block manager to remove blocks. | Reynold Xin | 2012-12-13 | 9 | -170/+361 |
| | | | | | | | |||||
| * | | | | | | SPARK-616: Logging dead workers in Web UI. | Patrick Wendell | 2012-12-17 | 5 | -3/+19 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch keeps track of which workers have died and marks them as such in the master web UI. It also handles workers which die and re-register using different actor ID's. | ||||
| * | | | | | | Merge pull request #332 from JoshRosen/spark-607 | Matei Zaharia | 2012-12-14 | 1 | -12/+17 |
| |\ \ \ \ \ \ | | | | | | | | | | | | | | | | | Add try-finally to handle MapOutputTracker timeouts | ||||
| | * | | | | | | Add try-finally to handle MapOutputTracker timeouts. | Josh Rosen | 2012-12-13 | 1 | -12/+17 |
| | |/ / / / / | |||||
| * | | | | | | Code review cleanup. | Charles Reiss | 2012-12-13 | 1 | -1/+4 |
| | | | | | | | |||||
| * | | | | | | Have standalone cluster report exit codes to clients. Addresses SPARK-639. | Charles Reiss | 2012-12-13 | 8 | -24/+20 |
| | |_|_|/ / | |/| | | | | |||||
| * | | | | | Fixed the broken Java unit test from SPARK-635. | Reynold Xin | 2012-12-13 | 1 | -7/+9 |
| | |_|/ / | |/| | | | |||||
| * | | | | Merge pull request #329 from woggling/executor-status-codes | Matei Zaharia | 2012-12-13 | 8 | -15/+118 |
| |\ \ \ \ | | |_|/ / | |/| | | | Executor exit status codes | ||||
| | * | | | ExecutorLostReason -> ExecutorLossReason | Charles Reiss | 2012-12-13 | 4 | -6/+6 |
| | | | | | |||||
| | * | | | Extra whitespace in ExecutorExitCode | Charles Reiss | 2012-12-13 | 1 | -0/+3 |
| | | | | | |||||
| | * | | | Explain slaveLost calls made by StandaloneSchedulerBackend | Charles Reiss | 2012-12-13 | 1 | -5/+5 |
| | | | | | |||||
| | * | | | Log duplicate slaveLost() calls in ClusterScheduler. | Charles Reiss | 2012-12-13 | 1 | -1/+7 |
| | | | | | |||||
| | * | | | Normalize executor exit statuses and report them to the user. | Charles Reiss | 2012-12-13 | 8 | -11/+94 |
| | | | | | |||||
| | * | | | Call slaveLost on executor death for standalone clusters. | Charles Reiss | 2012-12-12 | 1 | -1/+12 |
| | |/ / | |||||
| * | | | SPARK-635: Pass a TaskContext object to compute() interface and use | Reynold Xin | 2012-12-13 | 20 | -49/+46 |
| | | | | | | | | | | | | | | | | that to close Hadoop input stream. Incorporated Matei's command. | ||||
| * | | | SPARK-635: Pass a TaskContext object to compute() interface and use that | Reynold Xin | 2012-12-13 | 26 | -205/+207 |
| |/ / | | | | | | | | | | to close Hadoop input stream. | ||||
| * | | Suppress shuffle block updates when a slave node comes back. | Reynold Xin | 2012-12-10 | 2 | -12/+17 |
| | | | |||||
| * | | Merge pull request #317 from woggling/block-manager-heartbeat | Matei Zaharia | 2012-12-10 | 6 | -62/+318 |
| |\ \ | | | | | | | | | Implement block manager heartbeat | ||||
| | * | | Decrease BlockManagerMaster logging verbosity | Charles Reiss | 2012-12-10 | 1 | -1/+2 |
| | | | | |||||
| | * | | Use Akka scheduler for BlockManager heart beats. | Charles Reiss | 2012-12-10 | 4 | -49/+41 |
| | | | | | | | | | | | | | | | | Adds required ActorSystem argument to BlockManager constructors. | ||||
| | * | | Changed default block manager heartbeat interval to 5 s | Charles Reiss | 2012-12-09 | 1 | -1/+1 |
| | | | | |||||
| | * | | Don't divide by milliseconds by 1000 more. | Charles Reiss | 2012-12-06 | 1 | -1/+1 |
| | | | | |||||
| | * | | map -> foreach | Charles Reiss | 2012-12-06 | 1 | -1/+1 |
| | | | | |||||
| | * | | Make LocalSparkCluster use distinct IPs | Charles Reiss | 2012-12-06 | 1 | -2/+6 |
| | | | | |||||
| | * | | Tests for block manager heartbeats. | Charles Reiss | 2012-12-05 | 1 | -0/+68 |
| | | | | |||||
| | * | | Add block manager heart beats. | Charles Reiss | 2012-12-05 | 2 | -25/+193 |
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Renames old message called 'HeartBeat' to 'BlockUpdate'. The BlockManager periodically sends a heart beat message to the master. If the manager is currently not registered. The master responds to the heart beat by indicating whether the BlockManager is currently registered with the master. Additionally, the master now also responds to block updates by indicating whether the BlockManager in question is registered. When the BlockManager detects (by heart beat or failed block update) that it stopped being registered, it reregisters and sends block updates for all its blocks. | ||||
| | * | | Track block managers by hostname; handle manager removal. | Charles Reiss | 2012-12-05 | 1 | -3/+27 |
| | | | |