aboutsummaryrefslogtreecommitdiff
path: root/core
Commit message (Collapse)AuthorAgeFilesLines
* [SPARK-5645] Added local read bytes/time to task metricsKay Ousterhout2015-02-1213-28/+125
| | | | | | | | | | | | | | | | | | | | | | | | | ksakellis I stumbled on your JIRA for this yesterday; I know it's assigned to you but I'd already done this for my own uses a while ago so thought I could help save you the work of doing it! Hopefully this doesn't duplicate any work you've already done. Here's a screenshot of what the UI looks like: ![image](https://cloud.githubusercontent.com/assets/1108612/6135352/c03e7276-b11c-11e4-8f11-c6aefe1f35b9.png) Based on a discussion with pwendell, I put the data read remotely in as an additional metric rather than showing it in brackets as you'd suggested, Kostas. The assumption here is that the average user doesn't care about the differentiation between local / remote data, so it's better not to pollute the UI. I also added data about the local read time, which I've found very helpful for debugging, but I didn't put it in the UI because I think it's probably something not a ton of people will need to use. With this change, the total read time and total write time shown in the UI will be equal, fixing a long-term source of user confusion: ![image](https://cloud.githubusercontent.com/assets/1108612/6135399/25f14490-b11d-11e4-8086-20be5f4002e6.png) Author: Kay Ousterhout <kayousterhout@gmail.com> Closes #4510 from kayousterhout/SPARK-5645 and squashes the following commits: 4a0182c [Kay Ousterhout] oops 5f5da1b [Kay Ousterhout] Small style fix 5da04cf [Kay Ousterhout] Addressed more comments from Kostas ba05149 [Kay Ousterhout] Remove parens a9dc685 [Kay Ousterhout] Kostas comment, test fix 33d2e2d [Kay Ousterhout] Merge remote-tracking branch 'upstream/master' into SPARK-5645 347e2cd [Kay Ousterhout] [SPARK-5645] Added local read bytes/time to task metrics
* [SPARK-5655] Don't chmod700 application files if running in YARNAndrew Rowson2015-02-121-8/+3
| | | | | | | | | | | | | [Was previously PR4507] As per SPARK-5655, recently committed code chmod 700s all application files created on the local fs by a spark executor. This is both unnecessary and broken on YARN, where files created in the nodemanager's working directory are already owned by the user running the job and the 'yarn' group. Group read permission is also needed for the auxiliary shuffle service to be able to read the files, as this is running as the 'yarn' user. Author: Andrew Rowson <github@growse.com> Closes #4509 from growse/master and squashes the following commits: 7ca993c [Andrew Rowson] Moved chmod700 functionality into Utils.getOrCreateLocalRootDirs f57ce6b [Andrew Rowson] [SPARK-5655] Don't chmod700 application files if running in a YARN container
* [SPARK-5733] Error Link in Pagination of HistroyPage when showing Incomplete ↵guliangliang2015-02-111-4/+7
| | | | | | | | | | | | | | Applications The links in pagination of HistroyPage is wrong when showing Incomplete Applications. If "2" is click on the following page "http://history-server:18080/?page=1&showIncomplete=true", it will go to "http://history-server:18080/?page=2" instead of "http://history-server:18080/?page=2&showIncomplete=true". Author: guliangliang <guliangliang@qiyi.com> Closes #4523 from marsishandsome/Spark5733 and squashes the following commits: 9d7b593 [guliangliang] [SPARK-5733] Error Link in Pagination of HistroyPage when showing Incomplete Applications
* [SPARK-5729] Potential NPE in standalone REST APIAndrew Or2015-02-101-1/+1
| | | | | | | | | | If the user specifies a bad REST URL, the server will throw an NPE instead of propagating the error back. This is because the default `ErrorServlet` has the wrong prefix. This is a one line fix. I am will add more comprehensive tests in a separate patch. Author: Andrew Or <andrew@databricks.com> Closes #4518 from andrewor14/rest-npe and squashes the following commits: 16b15bc [Andrew Or] Correct ErrorServlet context prefix
* [SPARK-4879] Use driver to coordinate Hadoop output committing for ↵mcheah2015-02-1012-23/+549
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | speculative tasks Previously, SparkHadoopWriter always committed its tasks without question. The problem is that when speculation is enabled sometimes this can result in multiple tasks committing their output to the same file. Even though an HDFS-writing task may be re-launched due to speculation, the original task is not killed and may eventually commit as well. This can cause strange race conditions where multiple tasks that commit interfere with each other, with the result being that some partition files are actually lost entirely. For more context on these kinds of scenarios, see SPARK-4879. In Hadoop MapReduce jobs, the application master is a central coordinator that authorizes whether or not any given task can commit. Before a task commits its output, it queries the application master as to whether or not such a commit is safe, and the application master does bookkeeping as tasks are requesting commits. Duplicate tasks that would write to files that were already written to from other tasks are prohibited from committing. This patch emulates that functionality - the crucial missing component was a central arbitrator, which is now a module called the OutputCommitCoordinator. The coordinator lives on the driver and the executors can obtain a reference to this actor and request its permission to commit. As tasks commit and are reported as completed successfully or unsuccessfully by the DAGScheduler, the commit coordinator is informed of the task completion events as well to update its internal state. Future work includes more rigorous unit testing and extra optimizations should this patch cause a performance regression. It is unclear what the overall cost of communicating back to the driver on every hadoop-committing task will be. It's also important for those hitting this issue to backport this onto previous version of Spark because the bug has serious consequences, that is, data is lost. Currently, the OutputCommitCoordinator is only used when `spark.speculation` is true. It can be disabled by setting `spark.hadoop.outputCommitCoordination.enabled=false` in SparkConf. This patch is an updated version of #4155 (by mccheah), which in turn was an updated version of this PR. Closes #4155. Author: mcheah <mcheah@palantir.com> Author: Josh Rosen <joshrosen@databricks.com> Closes #4066 from JoshRosen/SPARK-4879-sparkhadoopwriter-fix and squashes the following commits: 658116b [Josh Rosen] Merge remote-tracking branch 'origin/master' into SPARK-4879-sparkhadoopwriter-fix ed783b2 [Josh Rosen] Address Andrew’s feedback. e7be65a [Josh Rosen] Merge remote-tracking branch 'origin/master' into SPARK-4879-sparkhadoopwriter-fix 14861ea [Josh Rosen] splitID -> partitionID in a few places ed8b554 [Josh Rosen] Merge remote-tracking branch 'origin/master' into SPARK-4879-sparkhadoopwriter-fix 48d5c1c [Josh Rosen] Roll back copiesRunning change in TaskSetManager 3969f5f [Josh Rosen] Re-enable guarding of commit coordination with spark.speculation setting. ede7590 [Josh Rosen] Add test to ensure that a job that denies all commits cannot complete successfully. 97da5fe [Josh Rosen] Use actor only for RPC; call methods directly in DAGScheduler. f582574 [Josh Rosen] Some cleanup in OutputCommitCoordinatorSuite a7c0e29 [Josh Rosen] Create fake TaskInfo using dummy fields instead of Mockito. 997b41b [Josh Rosen] Roll back unnecessary DAGSchedulerSingleThreadedProcessLoop refactoring: 459310a [Josh Rosen] Roll back TaskSetManager changes that broke other tests. dd00b7c [Josh Rosen] Move CommitDeniedException to executors package; remove `@DeveloperAPI` annotation. c79df98 [Josh Rosen] Some misc. code style + doc changes: f7d69c5 [Josh Rosen] Merge remote-tracking branch 'origin/master' into SPARK-4879-sparkhadoopwriter-fix 92e6dc9 [Josh Rosen] Bug fix: use task ID instead of StageID to index into authorizedCommitters. b344bad [Josh Rosen] (Temporarily) re-enable “always coordinate” for testing purposes. 0aec91e [Josh Rosen] Only coordinate when speculation is enabled; add configuration option to bypass new coordination. 594e41a [mcheah] Fixing a scalastyle error 60a47f4 [mcheah] Writing proper unit test for OutputCommitCoordinator and fixing bugs. d63f63f [mcheah] Fixing compiler error 9fe6495 [mcheah] Fixing scalastyle 1df2a91 [mcheah] Throwing exception if SparkHadoopWriter commit denied d431144 [mcheah] Using more concurrency to process OutputCommitCoordinator requests. c334255 [mcheah] Properly handling messages that could be sent after actor shutdown. 8d5a091 [mcheah] Was mistakenly serializing the accumulator in test suite. 9c6a4fa [mcheah] More OutputCommitCoordinator cleanup on stop() 78eb1b5 [mcheah] Better OutputCommitCoordinatorActor stopping; simpler canCommit 83de900 [mcheah] Making the OutputCommitCoordinatorMessage serializable abc7db4 [mcheah] TaskInfo can't be null in DAGSchedulerSuite f135a8e [mcheah] Moving the output commit coordinator from class into method. 1c2b219 [mcheah] Renaming oudated names for test function classes 66a71cd [mcheah] Removing whitespace modifications 6b543ba [mcheah] Removing redundant accumulator in unit test c9decc6 [mcheah] Scalastyle fixes bc80770 [mcheah] Unit tests for OutputCommitCoordinator 6e6f748 [mcheah] [SPARK-4879] Use the Spark driver to authorize Hadoop commits.
* [SPARK-5493] [core] Add option to impersonate user.Marcelo Vanzin2015-02-106-33/+79
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Hadoop has a feature that allows users to impersonate other users when submitting applications or talking to HDFS, for example. These impersonated users are referred generally as "proxy users". Services such as Oozie or Hive use this feature to run applications as the requesting user. This change makes SparkSubmit accept a new command line option to run the application as a proxy user. It also fixes the plumbing of the user name through the UI (and a couple of other places) to refer to the correct user running the application, which can be different than `sys.props("user.name")` even without proxies (e.g. when using kerberos). Author: Marcelo Vanzin <vanzin@cloudera.com> Closes #4405 from vanzin/SPARK-5493 and squashes the following commits: df82427 [Marcelo Vanzin] Clarify the reason for the special exception handling. 05bfc08 [Marcelo Vanzin] Remove unneeded annotation. 4840de9 [Marcelo Vanzin] Review feedback. 8af06ff [Marcelo Vanzin] Fix usage string. 2e4fa8f [Marcelo Vanzin] Merge branch 'master' into SPARK-5493 b6c947d [Marcelo Vanzin] Merge branch 'master' into SPARK-5493 0540d38 [Marcelo Vanzin] [SPARK-5493] [core] Add option to impersonate user.
* [SPARK-5644] [Core]Delete tmp dir when sc is stopSephiroth-Lin2015-02-102-1/+37
| | | | | | | | | | | | | | | | | | | | | | When we run driver as a service, and for each time we run job we only call sc.stop, then will not delete tmp dir create by HttpFileServer and SparkEnv, it will be deleted until the service process exit, so we need to delete these tmp dirs when sc is stop directly. Author: Sephiroth-Lin <linwzhong@gmail.com> Closes #4412 from Sephiroth-Lin/bug-fix-master-01 and squashes the following commits: fbbc785 [Sephiroth-Lin] using an interpolated string b968e14 [Sephiroth-Lin] using an interpolated string 4edf394 [Sephiroth-Lin] rename the variable and update comment 1339c96 [Sephiroth-Lin] add a member to store the reference of tmp dir b2018a5 [Sephiroth-Lin] check sparkFilesDir before delete f48a3c6 [Sephiroth-Lin] don't check sparkFilesDir, check executorId dd9686e [Sephiroth-Lin] format code b38e0f0 [Sephiroth-Lin] add dir check before delete d7ccc64 [Sephiroth-Lin] Change log level 1d70926 [Sephiroth-Lin] update comment e2a2b1b [Sephiroth-Lin] update comment aeac518 [Sephiroth-Lin] Delete tmp dir when sc is stop c0d5b28 [Sephiroth-Lin] Delete tmp dir when sc is stop
* [HOTFIX][SPARK-4136] Fix compilation and testsAndrew Or2015-02-103-11/+9
|
* SPARK-4136. Under dynamic allocation, cancel outstanding executor requests ↵Sandy Ryza2015-02-105-50/+184
| | | | | | | | | | | | | | | | | | | when no longer needed This takes advantage of the changes made in SPARK-4337 to cancel pending requests to YARN when they are no longer needed. Each time the timer in `ExecutorAllocationManager` strikes, we compute `maxNumNeededExecutors`, the maximum number of executors we could fill with the current load. This is calculated as the total number of running and pending tasks divided by the number of cores per executor. If `maxNumNeededExecutors` is below the total number of running and pending executors, we call `requestTotalExecutors(maxNumNeededExecutors)` to let the cluster manager know that it should cancel any pending requests above this amount. If not, `maxNumNeededExecutors` is just used as a bound in alongside the configured `maxExecutors` to limit the number of new requests. The patch modifies the API exposed by `ExecutorAllocationClient` for requesting additional executors by moving from `requestExecutors` to `requestTotalExecutors`. This makes the communication between the `ExecutorAllocationManager` and the `YarnAllocator` easier to reason about and removes some state that needed to be kept in the `CoarseGrainedSchedulerBackend`. I think an argument can be made that this makes for a less attractive user-facing API in `SparkContext`, but I'm having trouble envisioning situations where a user would want to use either of these APIs. This will likely break some tests, but I wanted to get feedback on the approach before adding tests and polishing. Author: Sandy Ryza <sandy@cloudera.com> Closes #4168 from sryza/sandy-spark-4136 and squashes the following commits: 37ce77d [Sandy Ryza] Warn on negative number cd3b2ff [Sandy Ryza] SPARK-4136
* SPARK-5239 [CORE] JdbcRDD throws "java.lang.AbstractMethodError: ↵Sean Owen2015-02-101-3/+3
| | | | | | | | | | | | oracle.jdbc.driver.xxxxxx.isClosed()Z" This is a completion of https://github.com/apache/spark/pull/4033 which was withdrawn for some reason. Author: Sean Owen <sowen@cloudera.com> Closes #4470 from srowen/SPARK-5239.2 and squashes the following commits: 2398bde [Sean Owen] Avoid use of JDBC4-only isClosed()
* [SPARK-5701] Only set ShuffleReadMetrics when task has shuffle depsKay Ousterhout2015-02-092-10/+40
| | | | | | | | | | The updateShuffleReadMetrics method in TaskMetrics (called by the executor heartbeater) will currently always add a ShuffleReadMetrics to TaskMetrics (with values set to 0), even when the task didn't read any shuffle data. ShuffleReadMetrics should only be added if the task reads shuffle data. Author: Kay Ousterhout <kayousterhout@gmail.com> Closes #4488 from kayousterhout/SPARK-5701 and squashes the following commits: 673ed58 [Kay Ousterhout] SPARK-5701: Only set ShuffleReadMetrics when task has shuffle deps
* [SPARK-5703] AllJobsPage throws empty.max exceptionAndrew Or2015-02-091-1/+3
| | | | | | | | | | | | If you have a `SparkListenerJobEnd` event without the corresponding `SparkListenerJobStart` event, then `JobProgressListener` will create an empty `JobUIData` with an empty `stageIds` list. However, later in `AllJobsPage` we call `stageIds.max`. If this is empty, it will throw an exception. This crashed my history server. Author: Andrew Or <andrew@databricks.com> Closes #4490 from andrewor14/jobs-page-max and squashes the following commits: 21797d3 [Andrew Or] Check nonEmpty before calling max
* [SPARK-2996] Implement userClassPathFirst for driver, yarn.Marcelo Vanzin2015-02-0918-161/+403
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Yarn's config option `spark.yarn.user.classpath.first` does not work the same way as `spark.files.userClassPathFirst`; Yarn's version is a lot more dangerous, in that it modifies the system classpath, instead of restricting the changes to the user's class loader. So this change implements the behavior of the latter for Yarn, and deprecates the more dangerous choice. To be able to achieve feature-parity, I also implemented the option for drivers (the existing option only applies to executors). So now there are two options, each controlling whether to apply userClassPathFirst to the driver or executors. The old option was deprecated, and aliased to the new one (`spark.executor.userClassPathFirst`). The existing "child-first" class loader also had to be fixed. It didn't handle resources, and it was also doing some things that ended up causing JVM errors depending on how things were being called. Author: Marcelo Vanzin <vanzin@cloudera.com> Closes #3233 from vanzin/SPARK-2996 and squashes the following commits: 9cf9cf1 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 a1499e2 [Marcelo Vanzin] Remove SPARK_HOME propagation. fa7df88 [Marcelo Vanzin] Remove 'test.resource' file, create it dynamically. a8c69f1 [Marcelo Vanzin] Review feedback. cabf962 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 a1b8d7e [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 3f768e3 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 2ce3c7a [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 0e6d6be [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 70d4044 [Marcelo Vanzin] Fix pyspark/yarn-cluster test. 0fe7777 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 0e6ef19 [Marcelo Vanzin] Move class loaders around and make names more meaninful. fe970a7 [Marcelo Vanzin] Review feedback. 25d4fed [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 3cb6498 [Marcelo Vanzin] Call the right loadClass() method on the parent. fbb8ab5 [Marcelo Vanzin] Add locking in loadClass() to avoid deadlocks. 2e6c4b7 [Marcelo Vanzin] Mention new setting in documentation. b6497f9 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 a10f379 [Marcelo Vanzin] Some feedback. 3730151 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 f513871 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 44010b6 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 7b57cba [Marcelo Vanzin] Remove now outdated message. 5304d64 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 35949c8 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 54e1a98 [Marcelo Vanzin] Merge branch 'master' into SPARK-2996 d1273b2 [Marcelo Vanzin] Add test file to rat exclude. fa1aafa [Marcelo Vanzin] Remove write check on user jars. 89d8072 [Marcelo Vanzin] Cleanups. a963ea3 [Marcelo Vanzin] Implement spark.driver.userClassPathFirst for standalone cluster mode. 50afa5f [Marcelo Vanzin] Fix Yarn executor command line. 7d14397 [Marcelo Vanzin] Register user jars in executor up front. 7f8603c [Marcelo Vanzin] Fix yarn-cluster mode without userClassPathFirst. 20373f5 [Marcelo Vanzin] Fix ClientBaseSuite. 55c88fa [Marcelo Vanzin] Run all Yarn integration tests via spark-submit. 0b64d92 [Marcelo Vanzin] Add deprecation warning to yarn option. 4a84d87 [Marcelo Vanzin] Fix the child-first class loader. d0394b8 [Marcelo Vanzin] Add "deprecated configs" to SparkConf. 46d8cf2 [Marcelo Vanzin] Update doc with new option, change name to "userClassPathFirst". a314f2d [Marcelo Vanzin] Enable driver class path isolation in SparkSubmit. 91f7e54 [Marcelo Vanzin] [yarn] Enable executor class path isolation. a853e74 [Marcelo Vanzin] Re-work CoarseGrainedExecutorBackend command line arguments. 89522ef [Marcelo Vanzin] Add class path isolation support for Yarn cluster mode.
* Add a config option to print DAG.KaiXinXiaoLei2015-02-091-0/+3
| | | | | | | | | | | | | | Add a config option "spark.rddDebug.enable" to check whether to print DAG info. When "spark.rddDebug.enable" is true, it will print information about DAG in the log. Author: KaiXinXiaoLei <huleilei1@huawei.com> Closes #4257 from KaiXinXiaoLei/DAGprint and squashes the following commits: d9fe42e [KaiXinXiaoLei] change log info c27ee76 [KaiXinXiaoLei] change log info 83c2b32 [KaiXinXiaoLei] change config option adcb14f [KaiXinXiaoLei] change the file. f4e7b9e [KaiXinXiaoLei] add a option to print DAG
* [SPARK-5698] Do not let user request negative # of executorsAndrew Or2015-02-091-0/+5
| | | | | | | | | | | | | Otherwise we might crash the ApplicationMaster. Why? Please see https://issues.apache.org/jira/browse/SPARK-5698. sryza I believe this is also relevant in your patch #4168. Author: Andrew Or <andrew@databricks.com> Closes #4483 from andrewor14/da-negative and squashes the following commits: 53ed955 [Andrew Or] Throw IllegalArgumentException instead 0e89fd5 [Andrew Or] Check against negative requests
* [SPARK-5691] Fixing wrong data structure lookup for dupe app registratio...mcheah2015-02-091-1/+1
| | | | | | | | | | | | | In Master's registerApplication method, it checks if the application had already registered by examining the addressToWorker hash map. In reality, it should refer to the addressToApp data structure, as this is what really tracks which apps have been registered. Author: mcheah <mcheah@palantir.com> Closes #4477 from mccheah/spark-5691 and squashes the following commits: efdc573 [mcheah] [SPARK-5691] Fixing wrong data structure lookup for dupe app registration
* [SPARK-5672][Web UI] Don't return `ERROR 500` when have missing argsKirill A. Korinskiy2015-02-086-16/+35
| | | | | | | | | | Spark web UI return `HTTP ERROR 500` when GET arguments is missing. Author: Kirill A. Korinskiy <catap@catap.ru> Closes #4239 from catap/ui_500 and squashes the following commits: 520e180 [Kirill A. Korinskiy] [SPARK-5672][Web UI] Return `HTTP ERROR 400` when have missing args
* SPARK-5408: Use -XX:MaxPermSize specified by user instead of default in ...Jacek Lewandowski2015-02-071-2/+12
| | | | | | | | | | ...ExecutorRunner and DriverRunner Author: Jacek Lewandowski <lewandowski.jacek@gmail.com> Closes #4203 from jacek-lewandowski/SPARK-5408-1.3 and squashes the following commits: d913686 [Jacek Lewandowski] SPARK-5408: Use -XX:MaxPermSize specified by used instead of default in ExecutorRunner and DriverRunner
* [SPARK-5388] Provide a stable application submission gateway for standalone ↵Andrew Or2015-02-0622-94/+2016
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | cluster mode The goal is to provide a stable, REST-based application submission gateway that is not inherently based on Akka, which is unstable across versions. This PR targets standalone cluster mode, but is implemented in a general enough manner that can be potentially extended to other modes in the future. Client mode is currently not included in the changes here because there are many more Akka messages exchanged there. As of the changes here, the Master will advertise two ports, 7077 and 6066. We need to keep around the old one (7077) for client mode and older versions of Spark submit. However, all new versions of Spark submit will use the REST gateway (6066). By the way this includes ~700 lines of tests and ~200 lines of license. Author: Andrew Or <andrew@databricks.com> Closes #4216 from andrewor14/rest and squashes the following commits: 8d7ce07 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest 6f0c597 [Andrew Or] Use nullable fields for integer and boolean values dfe4bd7 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest b9e2a08 [Andrew Or] Minor comments 02b5cea [Andrew Or] Fix tests d2b1ef8 [Andrew Or] Comment changes + minor code refactoring across the board 9c82a36 [Andrew Or] Minor comment and wording updates b4695e7 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest c9a8ad7 [Andrew Or] Do not include appResource and mainClass as properties 6fc7670 [Andrew Or] Report REST server response back to the user 40e6095 [Andrew Or] Pass submit parameters through system properties cbd670b [Andrew Or] Include unknown fields, if any, in server response 9fee16f [Andrew Or] Include server protocol version on mismatch 09f873a [Andrew Or] Fix style 8188e61 [Andrew Or] Upgrade Jackson from 2.3.0 to 2.4.4 37538e0 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest 9165ae8 [Andrew Or] Fall back to Akka if endpoint was not REST 252d53c [Andrew Or] Clean up server error handling behavior further c643f64 [Andrew Or] Fix style bbbd329 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest 792e112 [Andrew Or] Use specific HTTP response codes on error f98660b [Andrew Or] Version the protocol and include it in REST URL 721819f [Andrew Or] Provide more REST-like interface for submit/kill/status 581f7bf [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest 9e0d1af [Andrew Or] Move some classes around to reduce number of files (minor) 42e5de4 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest 1f1c03f [Andrew Or] Use Jackson's DefaultScalaModule to simplify messages 9229433 [Andrew Or] Reduce duplicate naming in REST field ade28fd [Andrew Or] Clean up REST response output in Spark submit b2fef8b [Andrew Or] Abstract the success field to the general response 6c57b4b [Andrew Or] Increase timeout in end-to-end tests bf696ff [Andrew Or] Add checks for enabling REST when using kill/status 7ee6737 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest e2f7f5f [Andrew Or] Provide more safeguard against missing fields 9581df7 [Andrew Or] Clean up uses of exceptions 914fdff [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest e2104e6 [Andrew Or] stable -> rest 3db7379 [Andrew Or] Fix comments and name fields for better error messages 8d43486 [Andrew Or] Replace SubmitRestProtocolAction with class name df90e8b [Andrew Or] Use Jackson for JSON de/serialization d7a1f9f [Andrew Or] Fix local cluster tests efa5e18 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest e42c131 [Andrew Or] Add end-to-end tests for standalone REST protocol 837475b [Andrew Or] Show the REST port on the Master UI d8d3717 [Andrew Or] Use a daemon thread pool for REST server 6568ca5 [Andrew Or] Merge branch 'master' of github.com:apache/spark into rest 77774ba [Andrew Or] Minor fixes 206cae4 [Andrew Or] Refactor and add tests for the REST protocol 63c05b3 [Andrew Or] Remove MASTER as a field (minor) 9e21b72 [Andrew Or] Action -> SparkSubmitAction (minor) 51c5ca6 [Andrew Or] Distinguish client and server side Spark versions b44e103 [Andrew Or] Implement status requests + fix validation behavior 120ab9d [Andrew Or] Support kill and request driver status through SparkSubmit 544de1d [Andrew Or] Major clean ups in code and comments e958cae [Andrew Or] Supported nested values in messages 484bd21 [Andrew Or] Specify an ordering for fields in SubmitDriverRequestMessage 6ff088d [Andrew Or] Rename classes to generalize REST protocol af9d9cb [Andrew Or] Integrate REST protocol in standalone mode 53e7c0e [Andrew Or] Initial client, server, and all the messages
* [SPARK-4874] [CORE] Collect record count metricsKostas Sakellis2015-02-0620-146/+548
| | | | | | | | | | | | | | | | | | | | | | | | Collects record counts for both Input/Output and Shuffle Metrics. For the input/output metrics, it just appends the counter every time the iterators get accessed. For shuffle on the write side, we count the metrics post aggregation (after a map side combine) and on the read side we count the metrics pre aggregation. This allows both the bytes read/written metrics and the records read/written to line up. For backwards compatibility, if we deserialize an older event that doesn't have record metrics, we set the metric to -1. Author: Kostas Sakellis <kostas@cloudera.com> Closes #4067 from ksakellis/kostas-spark-4874 and squashes the following commits: bd919be [Kostas Sakellis] Changed 'Records Read' in shuffleReadMetrics json output to 'Total Records Read' dad4d57 [Kostas Sakellis] Add a comment and check to BlockObjectWriter so that it cannot be reopend. 6f236a1 [Kostas Sakellis] Renamed _recordsWritten in ShuffleWriteMetrics to be more consistent 70620a0 [Kostas Sakellis] CR Feedback 17faa3a [Kostas Sakellis] Removed AtomicLong in favour of using Long b6f9923 [Kostas Sakellis] Merge AfterNextInterceptingIterator with InterruptableIterator to save a function call 46c8186 [Kostas Sakellis] Combined Bytes and # records into one column 57551c1 [Kostas Sakellis] Conforms to SPARK-3288 6cdb44e [Kostas Sakellis] Removed the generic InterceptingIterator and repalced it with specific implementation 1aa273c [Kostas Sakellis] CR Feedback 1bb78b1 [Kostas Sakellis] [SPARK-4874] [CORE] Collect record count metrics
* [SPARK-5600] [core] Clean up FsHistoryProvider test, fix app sort order.Marcelo Vanzin2015-02-062-48/+52
| | | | | | | | | | | | | | | Clean up some test setup code to remove duplicate instantiation of the provider. Also make sure unfinished apps are sorted correctly. Author: Marcelo Vanzin <vanzin@cloudera.com> Closes #4370 from vanzin/SPARK-5600 and squashes the following commits: 0d048d5 [Marcelo Vanzin] Cleanup test code a bit. 2585119 [Marcelo Vanzin] Review feedback. 8b97544 [Marcelo Vanzin] Merge branch 'master' into SPARK-5600 be979e9 [Marcelo Vanzin] Merge branch 'master' into SPARK-5600 298371c [Marcelo Vanzin] [SPARK-5600] [core] Clean up FsHistoryProvider test, fix app sort order.
* [SPARK-4361][Doc] Add more docs for Hadoop Configurationzsxwing2015-02-062-2/+46
| | | | | | | | | | I'm trying to point out reusing a Configuration in these APIs is dangerous. Any better idea? Author: zsxwing <zsxwing@gmail.com> Closes #3225 from zsxwing/SPARK-4361 and squashes the following commits: fe4e3d5 [zsxwing] Add more docs for Hadoop Configuration
* [HOTFIX] Fix test build break in ExecutorAllocationManagerSuite.Josh Rosen2015-02-061-6/+6
| | | | | | This was caused because #3486 added a new field to ExecutorInfo and #4369 added new tests that created ExecutorInfos. These patches were merged in quick succession and were never tested together, hence the compilation error.
* [SPARK-5555] Enable UISeleniumSuite testsJosh Rosen2015-02-061-9/+20
| | | | | | | | | | | | This patch enables UISeleniumSuite, a set of tests for the Spark application web UI. These tests were previously disabled because they were slow, but I think we now have sufficient test time budget that the benefit of enabling them outweighs the time costs. Author: Josh Rosen <joshrosen@databricks.com> Closes #4334 from JoshRosen/enable-uiseleniumsuite and squashes the following commits: 4ab9477 [Josh Rosen] Use BeforeAndAfterAll to cleanup WebDriver 71efc72 [Josh Rosen] Update broken UISeleniumSuite tests; use random port #. a5ab595 [Josh Rosen] Enable UISeleniumSuite tests.
* SPARK-2450 Adds executor log links to Web UIKostas Sakellis2015-02-0616-25/+140
| | | | | | | | | | | | | | | | | | | | | | | | Adds links to stderr/stdout in the executor tab of the webUI for: 1) Standalone 2) Yarn client 3) Yarn cluster This tries to add the log url support in a general way so as to make it easy to add support for all the cluster managers. This is done by using environment variables to pass to the executor the log urls. The SPARK_LOG_URL_ prefix is used and so additional logs besides stderr/stdout can also be added. To propagate this information to the UI we use the onExecutorAdded spark listener event. Although this commit doesn't add log urls when running on a mesos cluster, it should be possible to add using the same mechanism. Author: Kostas Sakellis <kostas@cloudera.com> Author: Josh Rosen <joshrosen@databricks.com> Closes #3486 from ksakellis/kostas-spark-2450 and squashes the following commits: d190936 [Josh Rosen] Fix a few minor style / formatting nits. Reset listener after each test Don't null listener out at end of main(). 8673fe1 [Kostas Sakellis] CR feedback. Hide the log column if there are no logs available 5bf6952 [Kostas Sakellis] [SPARK-2450] [CORE] Adds exeuctor log links to Web UI
* [SPARK-5618][Spark Core][Minor] Optimise utility code.Makoto Fukuhara2015-02-061-2/+2
| | | | | | | | Author: Makoto Fukuhara <fukuo33@gmail.com> Closes #4396 from fukuo33/fix-unnecessary-regex and squashes the following commits: cd07fd6 [Makoto Fukuhara] fix unnecessary regex.
* [SPARK-5593][Core]Replace BlockManagerListener with ExecutorListener in ↵lianhuiwang2015-02-062-22/+19
| | | | | | | | | | | | | ExecutorAllocationListener More strictly, in ExecutorAllocationListener, we need to replace onBlockManagerAdded, onBlockManagerRemoved with onExecutorAdded,onExecutorRemoved. because at some time, onExecutorAdded and onExecutorRemoved are more accurate to express these meanings. example at SPARK-5529, BlockManager has been removed,but executor is existed. andrewor14 sryza Author: lianhuiwang <lianhuiwang09@gmail.com> Closes #4369 from lianhuiwang/SPARK-5593 and squashes the following commits: 333367c [lianhuiwang] Replace BlockManagerListener with ExecutorListener in ExecutorAllocationListener
* [SPARK-4877] Allow user first classes to extend classes in the parent.Stephen Haberman2015-02-064-18/+61
| | | | | | | | | | | | | | | | | | | | Previously, the classloader isolation was almost too good, such that if a child class needed to load/reference a class that was only available in the parent, it could not do so. This adds tests for that case, the user-first Fake2 class extends the only-in-parent Fake3 class. It also sneaks in a fix where only the first stage seemed to work, and on subsequent stages, a LinkageError happened because classes from the user-first classpath were getting defined twice. Author: Stephen Haberman <stephen@exigencecorp.com> Closes #3725 from stephenh/4877_user_first_parent_inheritance and squashes the following commits: dabcd35 [Stephen Haberman] [SPARK-4877] Respect userClassPathFirst for the driver code too. 3d0fa7c [Stephen Haberman] [SPARK-4877] Allow user first classes to extend classes in the parent.
* [SPARK-5636] Ramp up faster in dynamic allocationAndrew Or2015-02-061-3/+3
| | | | | | | | | | A recent patch #4051 made the initial number default to 0. With this change, any Spark application using dynamic allocation's default settings will ramp up very slowly. Since we never request more executors than needed to saturate the pending tasks, it is safe to ramp up quickly. The current default of 60 may be too slow. Author: Andrew Or <andrew@databricks.com> Closes #4409 from andrewor14/dynamic-allocation-interval and squashes the following commits: d3cc485 [Andrew Or] Lower request interval
* [SPARK-5416] init Executor.threadPool before ExecutorSourceRyan Williams2015-02-061-3/+3
| | | | | | | | | | | Some ExecutorSource metrics can NPE by attempting to reference the threadpool otherwise. Author: Ryan Williams <ryan.blake.williams@gmail.com> Closes #4212 from ryan-williams/threadpool and squashes the following commits: 236f2ad [Ryan Williams] init Executor.threadPool before ExecutorSource
* [SPARK-5470][Core]use defaultClassLoader to load classes in KryoSerializerlianhuiwang2015-02-061-9/+2
| | | | | | | | | | | | | | Now KryoSerializer load classes of classesToRegister at the time of its initialization. when we set spark.kryo.classesToRegister=class1, it will throw SparkException("Failed to load class to register with Kryo". because in KryoSerializer's initialization, classLoader cannot include class of user's jars. we need to use defaultClassLoader of Serializer in newKryo(), because executor will reset defaultClassLoader of Serializer after Serializer's initialization. thank zzcclp for reporting it to me. Author: lianhuiwang <lianhuiwang09@gmail.com> Closes #4258 from lianhuiwang/SPARK-5470 and squashes the following commits: 73b719f [lianhuiwang] do the splitting and filtering during initialization 64cf306 [lianhuiwang] use defaultClassLoader to load classes of classesToRegister in KryoSerializer
* [SPARK-5582] [history] Ignore empty log directories.Marcelo Vanzin2015-02-062-8/+32
| | | | | | | | | | | | Empty log directories are not useful at the moment, but if one ends up showing in the log root, it breaks the code that checks for log directories. Author: Marcelo Vanzin <vanzin@cloudera.com> Closes #4352 from vanzin/SPARK-5582 and squashes the following commits: 1a6a3d4 [Marcelo Vanzin] [SPARK-5582] Fix exception when looking at empty directories.
* SPARK-5557: Explicitly include servlet API in dependencies.Patrick Wendell2015-02-051-0/+7
| | | | | | | | | | | | | Because of the way we shade jetty, we lose its dependency orbit in the assembly jar, which includes the javax servlet API's. This adds back orbit explicitly, using the version that matches our jetty version. Author: Patrick Wendell <patrick@databricks.com> Closes #4411 from pwendell/servlet-api and squashes the following commits: 445f868 [Patrick Wendell] SPARK-5557: Explicitly include servlet API in dependencies.
* [SPARK-5182] [SPARK-5528] [SPARK-5509] [SPARK-3575] [SQL] Parquet data ↵Cheng Lian2015-02-051-1/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | source improvements This PR adds three major improvements to Parquet data source: 1. Partition discovery While reading Parquet files resides in Hive style partition directories, `ParquetRelation2` automatically discovers partitioning information and infers partition column types. This is also a partial work for [SPARK-5182] [1], which aims to provide first class partitioning support for the data source API. Related code in this PR can be easily extracted to the data source API level in future versions. 1. Schema merging When enabled, Parquet data source collects schema information from all Parquet part-files and tries to merge them. Exceptions are thrown when incompatible schemas are detected. This feature is controlled by data source option `parquet.mergeSchema`, and is enabled by default. 1. Metastore Parquet table conversion moved to analysis phase This greatly simplifies the conversion logic. `ParquetConversion` strategy can be removed once the old Parquet implementation is removed in the future. This version of Parquet data source aims to entirely replace the old Parquet implementation. However, the old version hasn't been removed yet. Users can fall back to the old version by turning off SQL configuration `spark.sql.parquet.useDataSourceApi`. Other JIRA tickets fixed as side effects in this PR: - [SPARK-5509] [3]: `EqualTo` now uses a proper `Ordering` to compare binary types. - [SPARK-3575] [4]: Metastore schema is now preserved and passed to `ParquetRelation2` via data source option `parquet.metastoreSchema`. TODO: - [ ] More test cases for partition discovery - [x] Fix write path after data source write support (#4294) is merged It turned out to be non-trivial to fall back to old Parquet implementation on the write path when Parquet data source is enabled. Since we're planning to include data source write support in 1.3.0, I simply ignored two test cases involving Parquet insertion for now. - [ ] Fix outdated comments and documentations PS: This PR looks big, but more than a half of the changed lines in this PR are trivial changes to test cases. To test Parquet with and without the new data source, almost all Parquet test cases are moved into wrapper driver functions. This introduces hundreds of lines of changes. [1]: https://issues.apache.org/jira/browse/SPARK-5182 [2]: https://issues.apache.org/jira/browse/SPARK-5528 [3]: https://issues.apache.org/jira/browse/SPARK-5509 [4]: https://issues.apache.org/jira/browse/SPARK-3575 <!-- Reviewable:start --> [<img src="https://reviewable.io/review_button.png" height=40 alt="Review on Reviewable"/>](https://reviewable.io/reviews/apache/spark/4308) <!-- Reviewable:end --> Author: Cheng Lian <lian@databricks.com> Closes #4308 from liancheng/parquet-partition-discovery and squashes the following commits: b6946e6 [Cheng Lian] Fixes MiMA issues, addresses comments 8232e17 [Cheng Lian] Write support for Parquet data source a49bd28 [Cheng Lian] Fixes spelling typo in trait name "CreateableRelationProvider" 808380f [Cheng Lian] Fixes issues introduced while rebasing 50dd8d1 [Cheng Lian] Addresses @rxin's comment, fixes UDT schema merging adf2aae [Cheng Lian] Fixes compilation error introduced while rebasing 4e0175f [Cheng Lian] Fixes Python Parquet API, we need Py4J array to call varargs method 0d8ec1d [Cheng Lian] Adds more test cases b35c8c6 [Cheng Lian] Fixes some typos and outdated comments dd704fd [Cheng Lian] Fixes Python Parquet API 596c312 [Cheng Lian] Uses switch to control whether use Parquet data source or not 7d0f7a2 [Cheng Lian] Fixes Metastore Parquet table conversion a1896c7 [Cheng Lian] Fixes all existing Parquet test suites except for ParquetMetastoreSuite 5654c9d [Cheng Lian] Draft version of Parquet partition discovery and schema merging
* SPARK-5548: Fixed a race condition in AkkaUtilsSuiteJacek Lewandowski2015-02-051-1/+1
| | | | | | | | | | `Await.result` and `selection.resolveOne` runs the same timeout simultaneously. When `Await.result` timeout is reached first, then `TimeoutException` is thrown. On the other hand, when `selection.resolveOne` timeout is reached first, `ActorNotFoundException` is thrown. This is an obvious race condition and the easiest way to fix it is to increase the timeout of one method to make sure the code fails on the other method first. Author: Jacek Lewandowski <lewandowski.jacek@gmail.com> Closes #4343 from jacek-lewandowski/SPARK-5548-1.3 and squashes the following commits: b9ba47e [Jacek Lewandowski] SPARK-5548: Fixed a race condition in AkkaUtilsSuite
* SPARK-4687. Add a recursive option to the addFile APISandy Ryza2015-02-054-27/+237
| | | | | | | | | | | | | | | | | | | | | | | | | This adds a recursive option to the addFile API to satisfy Hive's needs. It only allows specifying HDFS dirs that will be copied down on every executor. There are a couple outstanding questions. * Should we allow specifying local dirs as well? The best way to do this would probably be to archive them. The drawback is that it would require a fair bit of code that I don't know of any current use cases for. * The addFiles implementation has a caching component that I don't entirely understand. What events are we caching between? AFAICT it's users calling addFile on the same file in the same app at different times? Do we want/need to add something similar for addDirectory. * The addFiles implementation will check to see if an added file already exists and has the same contents. I imagine we want the same behavior, so planning to add this unless people think otherwise. I plan to add some tests if people are OK with the approach. Author: Sandy Ryza <sandy@cloudera.com> Closes #3670 from sryza/sandy-spark-4687 and squashes the following commits: f9fc77f [Sandy Ryza] Josh's comments 70cd24d [Sandy Ryza] Add another test 13da824 [Sandy Ryza] Revert executor changes 38bf94d [Sandy Ryza] Marcelo's comments ca83849 [Sandy Ryza] Add addFile test 1941be3 [Sandy Ryza] Fix test and avoid HTTP server in local mode 31f15a9 [Sandy Ryza] Use cache recursively and fix some compile errors 0239c3d [Sandy Ryza] Change addDirectory to addFile with recursive 46fe70a [Sandy Ryza] SPARK-4687. Add a addDirectory API
* [SPARK-5411] Allow SparkListeners to be specified in SparkConf and loaded ↵Josh Rosen2015-02-043-21/+95
| | | | | | | | | | | | | | | | | | | | | | | | | | | when creating SparkContext This patch introduces a new configuration option, `spark.extraListeners`, that allows SparkListeners to be specified in SparkConf and registered before the SparkContext is initialized. From the configuration documentation: > A comma-separated list of classes that implement SparkListener; when initializing SparkContext, instances of these classes will be created and registered with Spark's listener bus. If a class has a single-argument constructor that accepts a SparkConf, that constructor will be called; otherwise, a zero-argument constructor will be called. If no valid constructor can be found, the SparkContext creation will fail with an exception. This motivation for this patch is to allow monitoring code to be easily injected into existing Spark programs without having to modify those programs' code. Author: Josh Rosen <joshrosen@databricks.com> Closes #4111 from JoshRosen/SPARK-5190-register-sparklistener-in-sc-constructor and squashes the following commits: 8370839 [Josh Rosen] Two minor fixes after merging with master 6e0122c [Josh Rosen] Merge remote-tracking branch 'origin/master' into SPARK-5190-register-sparklistener-in-sc-constructor 1a5b9a0 [Josh Rosen] Remove SPARK_EXTRA_LISTENERS environment variable. 2daff9b [Josh Rosen] Add a couple of explanatory comments for SPARK_EXTRA_LISTENERS. b9973da [Josh Rosen] Add test to ensure that conf and env var settings are merged, not overriden. d6f3113 [Josh Rosen] Use getConstructors() instead of try-catch to find right constructor. d0d276d [Josh Rosen] Move code into setupAndStartListenerBus() method b22b379 [Josh Rosen] Instantiate SparkListeners from classes listed in configurations. 9c0d8f1 [Josh Rosen] Revert "[SPARK-5190] Allow SparkListeners to be registered before SparkContext starts." 217ecc0 [Josh Rosen] Revert "Add addSparkListener to JavaSparkContext" 25988f3 [Josh Rosen] Add addSparkListener to JavaSparkContext 163ba19 [Josh Rosen] [SPARK-5190] Allow SparkListeners to be registered before SparkContext starts.
* [SPARK-4939] move to next locality when no pending tasksDavies Liu2015-02-042-8/+101
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently, if there are different locality in a task set, the tasks with NODE_LOCAL only get scheduled after all the PROCESS_LOCAL tasks are scheduled and timeout with spark.locality.wait.process (3 seconds by default). In local mode, the LocalScheduler will never call resourceOffer() again once it failed to get a task with same locality, then all the NODE_LOCAL tasks will be never scheduled. This bug could be reproduced by run example python/streaming/stateful_network_wordcount.py, it will hang after finished a batch with some data. This patch will check whether there is task for current locality level, if not, it will change to next locality level without waiting for `spark.locality.wait.process` seconds. It works for all locality levels. Because the list of pending tasks are updated lazily, the check can be false-positive, it means it will not move to next locality level even there is no valid pending tasks, it will wait for timeout. Author: Davies Liu <davies@databricks.com> Closes #3779 from davies/local_streaming and squashes the following commits: 2d25fb3 [Davies Liu] Update TaskSetManager.scala 1550668 [Davies Liu] add comment 1c37aac [Davies Liu] address comments 6b13824 [Davies Liu] address comments 906f456 [Davies Liu] Merge branch 'master' of github.com:apache/spark into local_streaming 414e79e [Davies Liu] fix bug, add logging ff8eabb [Davies Liu] Merge branch 'master' into local_streaming 28d1b3c [Davies Liu] check tasks 9d0ceab [Davies Liu] Merge branch 'master' of github.com:apache/spark into local_streaming 37a2804 [Davies Liu] fix tests 49bda82 [Davies Liu] address comment d8fb95a [Davies Liu] move to next locality level if no more tasks 2d6ae73 [Davies Liu] add comments 32d363f [Davies Liu] add regression test 7d8c5a5 [Davies Liu] jump to next locality if no pending tasks for executors
* [SPARK-5588] [SQL] support select/filter by SQL expressionDavies Liu2015-02-041-1/+10
| | | | | | | | | | | | | | | ``` df.selectExpr('a + 1', 'abs(age)') df.filter('age > 3') df[ df.age > 3 ] df[ ['age', 'name'] ] ``` Author: Davies Liu <davies@databricks.com> Closes #4359 from davies/select_expr and squashes the following commits: d99856b [Davies Liu] support select/filter by SQL expression
* [SPARK-5574] use given name prefix in dirImran Rashid2015-02-041-1/+1
| | | | | | | | | | | | | https://issues.apache.org/jira/browse/SPARK-5574 very minor, doesn't effect external behavior at all. Note that after this change, some of these dirs no longer will have "spark" in the name at all. I could change those locations that do pass in a name prefix to also include "spark", eg. "blockmgr" -> "spark-blockmgr" Author: Imran Rashid <irashid@cloudera.com> Closes #4344 from squito/SPARK-5574 and squashes the following commits: 33a84fe [Imran Rashid] use given name prefix in dir
* [SPARK-5341] Use maven coordinates as dependencies in spark-shell and ↵Burak Yavuz2015-02-036-6/+398
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | spark-submit This PR adds support for using maven coordinates as dependencies to spark-shell. Coordinates can be provided as a comma-delimited string after the flag `--packages`. Additional remote repositories (like SonaType) can be supplied as a comma-delimited string after the flag `--repositories`. Uses the Ivy library to resolve dependencies. Unfortunately the library has no decent documentation, therefore solving more complex dependency issues can be a problem. pwendell, mateiz, mengxr **Note: This is still a WIP. The following need to be handled:** - [x] add docs for the methods - [x] take local ivy cache path as an argument - [x] add tests - [x] add Windows compatibility - [x] exclude unused Ivy dependencies Author: Burak Yavuz <brkyvz@gmail.com> Closes #4215 from brkyvz/SPARK-5341ivy and squashes the following commits: 9215851 [Burak Yavuz] ready to merge db2a5cc [Burak Yavuz] changed logging to printStream 9dae87f [Burak Yavuz] file separators changed 71c374d [Burak Yavuz] merge conflicts fixed c08dc9f [Burak Yavuz] fixed merge conflicts 3ada19a [Burak Yavuz] fixed Jenkins error (hopefully) and added comment on oro 43c2290 [Burak Yavuz] fixed that ONE line 231f72f [Burak Yavuz] addressed code review 2cd6562 [Burak Yavuz] Merge branch 'master' of github.com:apache/spark into SPARK-5341ivy 85ec5a3 [Burak Yavuz] added oro as a dependency explicitly ea44ca4 [Burak Yavuz] add oro back to dependencies cef0e24 [Burak Yavuz] IntelliJ is just messing things up 97c4a92 [Burak Yavuz] fix more weird IntelliJ formatting 9cf077d [Burak Yavuz] fix weird IntelliJ formatting dcf5e13 [Burak Yavuz] fix windows command line flags 3a23f21 [Burak Yavuz] excluded ivy dependencies 53423e0 [Burak Yavuz] tests added 3705907 [Burak Yavuz] remove ivy-repo as a command line argument. Use global ivy cache as default c04d885 [Burak Yavuz] take path to ivy cache as a conf 2edc9b5 [Burak Yavuz] managed to exclude Spark and it's dependencies a0870af [Burak Yavuz] add docs. remove unnecesary new lines 6645af4 [Burak Yavuz] [SPARK-5341] added base implementation 882c4c8 [Burak Yavuz] added maven dependency download
* [SPARK-4939] revive offers periodically in LocalBackendDavies Liu2015-02-031-1/+10
| | | | | | | | | | | | | | | | | | | | The locality timeout assume that the SchedulerBackend can revive offers periodically, but currently LocalBackend did do that, then some job with mixed locality levels in local mode will hang forever. This PR let LocalBackend revive offers periodically, just like in cluster mode. Author: Davies Liu <davies@databricks.com> Closes #4147 from davies/revive and squashes the following commits: 2acdf9d [Davies Liu] Update LocalBackend.scala 3c8ca7c [Davies Liu] Update LocalBackend.scala d1b60d2 [Davies Liu] address comments from Kay 33ac9bb [Davies Liu] fix build d0da0d5 [Davies Liu] Merge branch 'master' of github.com:apache/spark into revive 6cf5972 [Davies Liu] fix thread-safety ed62a31 [Davies Liu] fix scala style df9008b [Davies Liu] fix typo bfc1396 [Davies Liu] revive offers periodically in LocalBackend
* [SPARK-4969][STREAMING][PYTHON] Add binaryRecords to streamingfreeman2015-02-031-1/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In Spark 1.2 we added a `binaryRecords` input method for loading flat binary data. This format is useful for numerical array data, e.g. in scientific computing applications. This PR adds support for the same format in Streaming applications, where it is similarly useful, especially for streaming time series or sensor data. Summary of additions - adding `binaryRecordsStream` to Spark Streaming - exposing `binaryRecordsStream` in the new PySpark Streaming - new unit tests in Scala and Python This required adding an optional Hadoop configuration param to `fileStream` and `FileInputStream`, but was otherwise straightforward. tdas davies Author: freeman <the.freeman.lab@gmail.com> Closes #3803 from freeman-lab/streaming-binary-records and squashes the following commits: b676534 [freeman] Clarify note 5ff1b75 [freeman] Add note to java streaming context eba925c [freeman] Simplify notes c4237b8 [freeman] Add experimental tag 30eba67 [freeman] Add filter and newFilesOnly alongside conf c2cfa6d [freeman] Expose new version of fileStream with conf in java 34d20ef [freeman] Add experimental tag 14bca9a [freeman] Add experimental tag b85bffc [freeman] Formatting 47560f4 [freeman] Space formatting 9a3715a [freeman] Refactor to reflect changes to FileInputSuite 7373f73 [freeman] Add note and defensive assertion for byte length 3ceb684 [freeman] Merge remote-tracking branch 'upstream/master' into streaming-binary-records 317b6d1 [freeman] Make test inline fcb915c [freeman] Formatting becb344 [freeman] Formatting d3e75b2 [freeman] Add tests in python a4324a3 [freeman] Line length 029d49c [freeman] Formatting 1c739aa [freeman] Simpler default arg handling 94d90d0 [freeman] Spelling 2843e9d [freeman] Add params to docstring 8b70fbc [freeman] Reorganization 28bff9b [freeman] Fix missing arg 9398bcb [freeman] Expose optional hadoop configuration 23dd69f [freeman] Tests for binaryRecordsStream 36cb0fd [freeman] Add binaryRecordsStream to scala fe4e803 [freeman] Add binaryRecordStream to Java API ecef0eb [freeman] Add binaryRecordsStream to python 8550c26 [freeman] Expose additional argument combination
* [SPARK-4795][Core] Redesign the "primitive type => Writable" implicit APIs ↵zsxwing2015-02-037-32/+128
| | | | | | | | | | | | | | | | | | to make them be activated automatically Try to redesign the "primitive type => Writable" implicit APIs to make them be activated automatically and without breaking binary compatibility. However, this PR will breaking the source compatibility if people use `xxxToXxxWritable` occasionally. See the unit test in `graphx`. Author: zsxwing <zsxwing@gmail.com> Closes #3642 from zsxwing/SPARK-4795 and squashes the following commits: 914b2d6 [zsxwing] Add implicit back to the Writables methods 0b9017f [zsxwing] Add some docs a0e8509 [zsxwing] Merge branch 'master' into SPARK-4795 39343de [zsxwing] Fix the unit test 64853af [zsxwing] Reorganize the rest 'implicit' methods in SparkContext
* Minor: Fix TaskContext deprecated annotations.Reynold Xin2015-02-031-3/+3
| | | | | | | | | | Made a mistake in https://github.com/apache/spark/pull/4324 Author: Reynold Xin <rxin@databricks.com> Closes #4333 from rxin/taskcontext-deprecate and squashes the following commits: 61c44ee [Reynold Xin] Minor: Fix TaskContext deprecated annotations.
* [SPARK-5549] Define TaskContext interface in Scala.Reynold Xin2015-02-035-131/+183
| | | | | | | | | | | | So the interface documentation shows up in ScalaDoc. Author: Reynold Xin <rxin@databricks.com> Closes #4324 from rxin/TaskContext-scala and squashes the following commits: 2480a17 [Reynold Xin] comment 573756f [Reynold Xin] style fixes and javadoc fixes. 87dd537 [Reynold Xin] [SPARK-5549] Define TaskContext interface in Scala.
* [SPARK-5414] Add SparkFirehoseListener class for consuming all SparkListener ↵Josh Rosen2015-02-021-0/+115
| | | | | | | | | | | | | | | events There isn't a good way to write a SparkListener that receives all SparkListener events and which will be future-compatible (e.g. it will receive events introduced in newer versions of Spark without having to override new methods to process those events). To address this, this patch adds `SparkFirehoseListener`, a SparkListener implementation that receives all events and dispatches them to a single `onEvent` method (which can be overridden by users). Author: Josh Rosen <joshrosen@databricks.com> Closes #4210 from JoshRosen/firehose-listener and squashes the following commits: 223f579 [Josh Rosen] Expand comment to explain rationale for this being a Java class. ecdfaed [Josh Rosen] Add SparkFirehoseListener class for consuming all SparkListener events.
* [SPARK-3778] newAPIHadoopRDD doesn't properly pass credentials for secure hdfsThomas Graves2015-02-021-1/+6
| | | | | | | | | | | | | | | | | | | | .this was https://github.com/apache/spark/pull/2676 https://issues.apache.org/jira/browse/SPARK-3778 This affects if someone is trying to access secure hdfs something like: val lines = { val hconf = new Configuration() hconf.set("mapred.input.dir", "mydir") hconf.set("textinputformat.record.delimiter","\003432\n") sc.newAPIHadoopRDD(hconf, classOf[TextInputFormat], classOf[LongWritable], classOf[Text]) } Author: Thomas Graves <tgraves@apache.org> Closes #4292 from tgravescs/SPARK-3788 and squashes the following commits: cf3b453 [Thomas Graves] newAPIHadoopRDD doesn't properly pass credentials for secure hdfs on yarn
* [SPARK-5219][Core] Add locks to avoid scheduling race conditionszsxwing2015-02-022-3/+3
| | | | | | | | Author: zsxwing <zsxwing@gmail.com> Closes #4019 from zsxwing/SPARK-5219 and squashes the following commits: 36a8b4e [zsxwing] Add locks to avoid race conditions
* SPARK-3996: Add jetty servlet and continuations.Patrick Wendell2015-02-021-1/+11
| | | | | | | | | | | | These are needed transitively from the other Jetty libraries we include. It was not picked up by unit tests because we disable the UI. Author: Patrick Wendell <patrick@databricks.com> Closes #4323 from pwendell/jetty and squashes the following commits: d8669da [Patrick Wendell] SPARK-3996: Add jetty servlet and continuations.
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-19 04:43:30 +0000