- #1648 Support submit storm jar via dashboard
- Support StreamCQL over Gearpump
- #1924, fix storm test spec failure
- #1926, fix client side configuration does not work in CGroup
- #1648, support submit storm jar via rest interface
- #1933, Fixes some of the documentation typos, adds SBT configuration
- #1936, allow user to config kafka source start offset time
- #1777, Dynamic DAG failed after changing parallelism for many times
- #1932, Fix client config overwrites critical system config
- #1895, make GearpumpNimbus a standalone service
- #1910, enlarge timeout in MasterClient
- #1952, document how to use StreamCQL over Gearpump
- #1723, Add an experimental cgroup module for CPU isolation.
- #1767, Add docs and implementation to support storm 0.10.
- Make UI mobile phone friendly.
- #1919, fix AppMasterSpec
- #1723 add an experiment module to do cgroup CPU isolation.
- #1913, add warning for detected loop in DAG.
- #1896, move storm examples under integration test
- #1897, increase akka logger start up timeout for unit test
- #1843 Add a Rest service to launch a new worker
- #1904, fixed "DAG canvas context menu cannot be displayed."
- #1898: in integration test, there need a retry attempting for login
- #1882: extract the tool tip string to externalized file
- #1837, Use java.net.preferIPv4Stack to force IPv4
- #1885, Shorten the table view when disabled in mobile.
- #1883,URL redirection error when accessing index.html
- #1874, Format the login failure message to more readable format.
- #1876, Shell command yarnclient should use 1 as the the default count for "addWorker"
- #1877, Fix several navigation issues on Mobile phone
- #1830: removed cluster side menu and tuned main navbar
- #1681: show application last error in overview page
- #1866, use geardefault.conf to replace reference.conf to avoid confusion
- #1767, Add docs and implementation to support storm 0.10
- #1868: show correct data points when switch historical view and recent view.
- #1864, Add robots.txt to disable search engine crawling
- #1860, add licenses for included source and binary libraries.
- #1804, fix UI authentication document page format
- #1247, add authentication for UI
- #1855: when service is unavailable, do not query models without waiting.
- #1833 update performance related doc
- #1831, Services exits on bind failure
- #1779: show critical path latency for application
- #1733 dynamic dag fall back to last dag version in case of Master HA
- #1804, fix navbar on mobile
- #1713 regression on HBaseSink
- #1838 YARN failure trying to contact YarnAppMaster when launched in YARN
- #1816: Akka GC problem. After fixing, performance gain 30%
- #1820, http cache is not effective
- #1817 Problems launching Gearpump in YARN environment
- #1814: metrics data will not update
- #1812 add gearpump.verbose-gc in gear.conf
- #1754, use scala binary version in example jar name
- #1804, polish documents
- #1803 fix performance regression
- #1411. new UI experience. now the frontend UI can safely support applications with tens of thousands to tasks.
- #1514 Add Akka-stream 1.0 experimental module
- #1698, refactor on yarn experiment module, with bug fixes, and enhanced functions, like dynamically adding/removing machines.
- #1679, fix dynamic dag bug (fail to transit)
- #1547, add a SequenceFileSink
- #1761, add more source comments and fix several typo
- #1799, amend two PROXY typos in files
- #1411: new UI to allow to scale to thousands of tasks.
- #1794, split dispatchers of Task and Netty
- #1018 transfer Any in Message
- #1792, fix storm module config conflict
- #1785, [yarn] UI daemon process cannot be terminated
- #1789 fix DynamicDagSpec
- #1784, document the limitation of yarnclient command line.
- #1778, AppSubmitter "-config" option is not used
- #1782, add doc to run storm over Gearpump application on yarn
- #1774, remove MaxPermSize JVM option as JDK8 no longer support it.
- #1772, When upload application in UI, custom config file is not effective
- #1770, processor metrics is aggregated in wrong way
- #1768, TaskFilterAggregator is not working
- #1763, change all GearPump to Gearpump
- #1411: metrics charts would not update after 15 seconds
- #1756, add HistoryMetricsConfig for master info and worker info
- #1754 standardize archive name
- #1753 update server channel group name
- #1747: storm spec failure due to parse application id
- #1747: updated the way to parse application id in command-line
- #1743, fixing some typo in the example folder
- #1744, Separate global config and application specific config, allow to config log level for specific class.
- #1734, handle smoothly when more resources are allocated than required.
- #1698, refactor on gearpump-yarn experiment module
- #1716: enable code coverage check for master branch
- #1720, support automatic integration test
- #1719 unhandled PutKVSuccess in Master
- #1714: some navbar visual issues
- #1499, Aggregate the metrics on server
- #1711: layout is more responsive on small devices
- #1679 dynamic dag will fall back when replacing a processor failed
- #1707, support html test report in integration test
- #1670, regression, change taskActor back to use stashing.
- #1703: task metrics in bar chart were empty
- #1701: wrong dirty check of modify processor and more
- #1384: rest service url was not set correctly
- #1682: allow specify a transition time when changing parallelism
- #1691: dashboard layout and textual changes and update to dashing…
- #1640 YARN deployment - no easy way of obtaining master address…
- #1688, Reduce unimportant logs in travis UT console
- #1671, make default timeout shorter to be more UI responsive.
- #1685, Should not publish gearpump-integration-test to maven
- #1683, Allow metrics to be aggregated before sending to UI
- #1223: will mark application with a warning when stalling task de…
- #1670, #1653, fix dynamic dag bug
- #1659 add a rest api to terminate service process
- #1672: when worker is killed, its detail page will no longer be updated
- #1660 remove Try block,give debug info to user instead.
- #1666: will should application details (and load metrics afterwards)
- #1651, fix change processor's parallelism on dashboard
- #1536, fix storm at least once support
- #1655: added a test case for replacing a processor
- #1653 fix replacing a processor
- #1639 refine daemon's classpath
- #1652, submit app with no arguments from dashboard returns warning
- #1547 add a SequenceFileSink
- #1424: when application is terminated, related status indicators …
- #1645: parse appid in submission response better
- #1514, add an experimental module to support akka-stream 1.0
- #1643, fix UI timeout issue when metrics is disabled
- #1632 Remove duplicate of repository
- #1630, StormBoltOutputCollector skips reporting on ack disabled
- New end-to-end integration test, better test coverage (#1243 Thanks to contribution by Stanley, Tianlun, Huafeng, Kewei)
- Storm binary compatibility. (still has some limitation)
- New Document site(#1506).
- Secure YARN and secure HBase support(#1458)
- spray is replaced by akka-http (#1261)
- New serialization implementation(#1445 allow user to custom a serialization framework).
- #1627, refine on custom serialization doc.
- #1624, fix the worker report resource timeout settings.
- #1384: removed wrong page redirection
- #1243: provide an integration test suite
- #1607 fix NONE_SESSION when restart tasks
- #1609, fix storm message timestamp bugs
- #1579 Additional documentation changes
- #1536, support storm ack
- #1581, fix HadoopCheckpointStore
- #1506 Totally reorganize documentation
- #1538 fix yarn application classpath
- #1519 fix performance regression
- #1485 no more pending messages in ExpressTransport
- #1458, enable secured HBase
- #1496, fix StormSerializer Kryo buffer overflow
- #1478: click on location button will copy text to clipboard
- #1491, fix ContainerLaunchContextSpec
- #1489, remove terminated executor from ExecutorManager
- #1481, improve storm over gearpump perf
- #1479: add uptime field in service interface
- #1481, fix storm performance
- #1482, fix TaskScheduler over schedules when multiple executors down at once
- #1458 support launching Gearpump in secured Yarn cluster
- #1476 fix serialization in transport layer
- #1470: updated a broken dashboard dependency
- #1463, fix storm config
- #1465 fix DistributedShell
- #1462 pass user config to SerializerPool
- #1453: allows use DEL to remove processor/edge-
- #491: Added support for composing DAG from scratch
- #1449 fix copy file to local in LocalJarStoreService-
- #1387 decouple serializer for Message
- #1445 refactor serialization implementation
- #1343, support Storm emitDirect and DirectGrouping
- #809: static resources are now sent to client gzipped
- #1438: polished modify processor dialog
- #1423: show full error log in a popup
- #1423: show error message when uploading is failed and polished the dialog
- #1430: turn off illegal header warning
- #1428, add unidoc plugin which allow us to build the scaladoc for the whole project(include all sub-projects)
- #1228, add data source and data sink for java api
- #1228, further simplify the java interface
- #1422: removed if-modified-since header. it is not relevant and is handled very different by different web servers
- #1228, add java interface for Graph API
- #1411: chart will always fill data points for 15 minutes
- #1261, #1127, replace spray with akka-http
- #1373 add duplicated edge detecting
- #1412: add 'no-cache' header in ajax responses
- #1194 add graph cycle detection
- #491, add a rest api to acquire Gearpump's built-in partitioners
- #1405: now the number of executors is back
- #1371: a right way to show historical metrics (2) added committed…
- #1402, fix NoneGroupingPartitioner
- #1399: in application overview page the task count of executor wa…
- #1397, allow KafkaSink to receive Message of bytes
- #1395, cross publishSigned
- #1374 remove jars of yarn in executor's classpath
- #1266: should exclude dead edges
- #1238, adds BroadcastPartitioner
- #1381 test on travis out of memory
- #1379, support storm tick tuple
Highlight: new version UI dashboard.
- #1369: object path and metrics name are divided by ':' now
- #1369: fixed data initialization issue of echart
- #1364, add default value for BYTES_PER_MESSAGE of SOLStreamProducer
- #1361: able to show multiple tasks in chart (better not select mo…
- #1358: properties will be updated (2) moved roboto font and echar…
Highlight: new version UI dashboard.
- #606: ui v2
- #1352, return appmaster config
- #1344 fix Master HA bug
- Rename package name to io.gearpump
- #1341, Cannot submit application in UI if one master is killed and another master started
- Allow user to change the DAG on the fly by providing a new jar.
- Add hadoop storage for Transactional topology
- shade environment dependencies jars.
- Support scala 2.10 and scala 2.11
- #1339, metrics typo
- #1336, verbose log to console
- #1331, check whether we have duplicate metric before registering.
- #1330, Worker cannot init when setting HDFS as jarStore
- #1334, fix Storm LocalCluster dependencies
- #1324, tag storm tuple with timestamp
- #1246 isolate processors' classpath
- #1327, move project state and dsl to under streaming
- #924, fix storm wordcount example
- #1319 fix application HA not working
- #1315, fix clock stalling
- #1305 remove JarFileContainer, replace it with Path.
- #1313, fix 2.10 build failure
- #601, 1. add jvm metrics 2. add master and worker metrics.
- #1296, enable cross compile and publish
- #606, add more rest service for new UI(version 2)
- #1013, track task checkpoint time at ClockService
- #1301, fix state example tests
- #339, add Hadoop-compatible CheckpointStore
- #1278, Optimize the build script for scalajs.
- #1285, fix travis deploy build script
- #1241 shade akka-kryo-serialization
- #1278 split services into 2 projects for scala.js
- #1290, purify state dependencies
- #1289, construct HTable at cluster side
- #1288: upgraded to vis 4.7.0 and optimized redraw
- #1217, Remove unnecessary dependencies from project core
- #1249, separate classpath for different daemon tools
- #1254, Add error reporting service and UI
- #1251, refine the build package layout.
- #1253, Partitioner instance is wrongly shared by multiple tasks
- #1223: stalling status might not be updated in expected way
- #1257, return error information when rest api failed.
- #1258: support filter metrics
- #1245, Replace processor won't work for data source
- #1267, print services help into to console
- #1269: could not specify transit time (2) improved time picker select behavior
- #1241 shade guava, gs-collections and codahale-metric
- #1275, fix HealthChecker
- #1273, refine the metrics UI
- #1270, add flow control for metrics data
- #1281, make Processor ReferenceEqual=
- #1244 simplify Build.scala
- #1175 refactor dsl source and sink to remove hbase/kafka dependencies.
- #1154: allow user to submit app by UI
- #1165, fix random UT failures, avoid create new process in UT
- #1166, throw exception when submit application failed. fix #1151, reduce the memory footprint of each Task
- #1098, decouple KafkaStorage from KafkaSource
- #1189 remove unnecessary dependency
- #1192, fix potential message loss exception in ExpressTransport
- #1191 change Kryo's instantiator strategy, add fallback strategy when serialization fails.
- #1156: (1) ui supports replace processor (2) network graph would not be destroyed (3) changed dropdown menu look and feel (4) updated js libraries
- #1208 zookeeper.connect is being overridden to localhost:2181
- #1210, fix kafka examples
- Better YARN support, and error handling.
- UI allow user to submit a application directly
- Split framework lib with application lib, to reduce class path pollution(ongoing). (#1017)
- Exactly once message processing API (#6).
- Improved Data Connector with Kafka, and Hbase. (#1012)
- Dynamic DAG(#101). Which allow user to replace a computation processor(e.g. Change the parallelism, or change to upgraded implementation Task class)
- #1154: UI, allow user to submit application jar in UI
- #1162: UI Backend, Rest backend cannot resume connection with master after master restart
- #1159: memory leak in worker thread pool.
- #1157: refactor rest interface and UT
- #1151: reduce the memory footprint when there are thousands of tasks
- #1149: Shell tools printed too much detail on console
- #1146: actor hungry when worker use block-io to wait response from FileServer.
- #1088: move hbase code to external/
- #1140:pass app name to task
- #1017: Split daemon dependencies with core dependencies
- #1144: fix out of memory when trying to scale gearpump to 2000 task on a 2 core machine
- #995: Command line parser should be able to fall back to MAIN-CLASS definition in MANIFEST.IN when mainClass is not specified in command line options.
- #1136: fix a dead-lock when shutting down actor system.
- #101: feature dynamic dag, which allows user to change the running topology in the fly.
- #1115: design better interface for kafka module
- #1123: UI, able to restart a running application via dashboard
- #1037: When user create daemon logs like ui.log in the running, we should be able to recreate a new log file.
- #1106: cache opensource webfont lato (831kb in total) locally so that there is no error when no internet access
- #1097: appdag should adapt to window size automatically.
- #1096: (1) echart would not response resizing (2) metric selector did not work due to angularjs 1.4 breaking change (3) tested and updated other js libraries
- #1094: service unreachable message will be shown when health check fails
- #1066: AppSubmitter should not return 0 when error occurred
- #1056: fix yarn client, to be able to submit applications to master when running in YARN
- #1083: fix KafkaWordCount example
- #1080: [UI] uses moving average 1 minute instead of mean rate. As the meanrate changes much slower.
- #1017, split the lib directory into daemon, and lib. So that user application can have less dependencies.
- #1067. [YARN]Changing visibility of yarn resources
- #1064: UI. edge data was not constructed correctly
- #1060: Add default kafka message decoder
- #1012, add Source and Sink API
- #1044 print worker hostname in master log when registering
- #922: UI. get version by rest api
- #1025: master is binding to incorrect port when deployed on Yarn
- #964: #966. Use FSM to manage state in YarnApplicationMaster. Adding missing UT.
- #1010, Simplify the travis build so that it take less time.
Highlight: experimental transactional support
- #1006, revert the display clock to original version
- #1004, fix transactional state compile failure
- #6, add transaction api
- #902, delete kafka topic on close
- #826: upgraded to visjs 4.1.0-develop
- #924, show application name and status on "gear info"
- #981, add prompt for ClassNotFoundException
- #977, speedup the UT by avoiding creating multiple sub process JVMs.
- #974, Remove unnecessary MasterProxy creation in unit test
- #969 Tests for examples create a test harness per test
- #968, update Codecov upload method
- #960, Services Specs do not terminate ClockService
- #959, LOG in TimeoutScheduler causes JVM exit
- #935 improve the application clock
- #659, remove kafka integration test
- #947 optimize checkMessage performance in TaskActor
- #945, [NPE regression] Partitioner is null
- #941, allow user to define custom partitioner by REST
- #890: add cloudera manager integration support for gearpump
- #928 continue to improve the performance of taskActor.allowSendingMoreMessages
- #936 bumping up version of sbt-scoverage plugin
- #934, allow user to define uid for edge partitioner.
- #928 improve the allowSendingMsg performance
- #926: previous restcall should be dropped if the page is destroyed
- #729 remove argument '-master' in YARN service and documents.
- #759, fix storm connector bug due to unstable topology sort of DAG
- #775, fix netty config
- #778, log improvements
- #781 when launching lots of tasks, the application failed to transfer message across hosts
- #782, a) add wildcard match to return metrics(support glob char … and *), b) add diagnosis message if the clock stop advancing
- #786, Read user config from classpath, the appmaster/executor wil use the dynamic user config
- #773: skew chart will show relative skew
- #790, 1) return detail task data in appmaster REST. 2) bind executor id with executor system id
- #795 TaskScheduleImpl'bug when executor failed
- #799 Getting 2.10 cross build working
- #802 add process id and host name in executor log file
- #803: (1) websocket is by default not preferred (2) throughput should be added as sum not mean (3) changed input/output message to sink/source processor receive/send throughput
- #805, metrics rest service should return latest metrics received
- #684 - setting -Xmx for master and worker when running on Yarn This will prevent JVMs from growing above limits and get killed by Yarn
- #796: (1) added executor info (2) fixed skew chart issue for the first node
- #801, add config service for master and worker.
- #741 add a example transport use case
- #814, expose TaskActor.minClock through TaskContext
- #741 refine example
- #817 split examples jar into multiple jars
- #824, allow to use default partitioner when defining a DAG
- #829, add some handy operator like groupByKey, sum, for KV Stream
- #831: uses pagination control to speedup table rendering
- #816: use multi-select control to select tasks
- #840: task charts data were incorrect
- #844, expose upstream minclock
- #204, page rank demo code
- #846, support more anyVals in user config
- #843, Can't put custom user config in application.conf
- #849, set default hostname to 127.0.0.1 in UT
- #851, JVM not exited when there is exception due to akka create non-daemon threads
- #854, fix storm connector performance
- #856, Service launch failed
- #853, fix resource leak(thread not closed, process not killed in UT. Also increase the PermGen size to avoid Permgen OOM.
- #859, random UT fail due to akka bug, "akka cluster deadlock when initializing"
- #865, Change the default timeout setting in akka test expectMsg
- #871, Add explicit error log for kryo serialization exception
- #877: source node and sink node were not calculated correctly.
- #874, [TaskActor] task onStart should be called after the network transport layer is ready
- #879: split metrics into different views and changed tooltip control
- #881: diverse issues of skew charts and made tooltip nice
- #876: clock is updated every second-
- #885: wrong application clock in some case
- #53, rest interface to submit a dag by JSON representation
- #887, add a rest to get stalling tasks
- #801: added download links for configurations-
- #742 add a rest to get Gearpump version
- #898, Downgrade akka version from 2.3.9 to 2.3.6
- #900, Use gearpump.hostname by default=
- #719 add Kafka Source and HBase Sink for dsl
- #905: Upgrade sbt-pack from 0.6.8 to 0.6.9
- #602: dashboard will freeze when server is unreachable
- #907: calculate application clock update frequency for a 30 second time frame
- #919: vis.js's version was not updated
- #768, Serious performance degration of ui server on windows
- #765, improve the graph type inference so we don't need to set template argument type explicitly
- #716, Refine the user interface. improve the user interface, like Application, Processor
- #755, fix Build breakage,
- #729, remove argument '-master' in command line option, use gear.conf instead.
- #713, provide an option to read from beginning of a topic
- #746, support state clock for task so that we can retrieve the state timestamp
- #744: fix several metrics issues
- #638, Use Subscription to link two processor
- #662, [UI] added processor details tab
- #735, [yarn]Launch UI in the same container of Yarn AppMaster container.
- #732: [UI]add more charts and add more metrics to dag
- #731 rename TaskDescription to ProcessorDescription
- #708, fix storm connector config classpath
- #722: [UI]use different color and opacity for edges
- #708, allow user to pass in a customized storm config
- #709: [UI]npe when streaming dag is not initialized
- #704, DAG processor name is "undefined" if "description` field of TaskDescription is not defined
- #706, Remove example jars from gearpump built-in classpath.
- #701 add a HBase sink
- #28, add an experiment module to support a very basic flatmap dsl
- #666, add UI for stock crawler example
- #691: [UI] fix metrics
- #680, add a service in appmaster to support query of task actorRef
- #210, allow easier remote debugging executor process
- #666, add a stock index crawler example,
- #658 make config of Gearpump on Yarn configurable
- #676, add a storm connector to allow user to run arbitrary storm jar
- #657, remove unnecessary yarn deps
- #672, extend Task to support unmanaged messages so that every task can serve as full functional service
- #670, fix yarn application path and log path
- #665, Add scheduleOnce interface in TaskContext
Change List:
- #654, Use yarn to distribute whole gearpump package instead of jars
- #631, remove unnecessary storm dependencies
- #652, log conflict between slf4j and log4j
- #652, log is muted when running on yarn
- #650, remove logback classic slf4j binding from classpath
- #648, yarn unable to start worker
- #27, Integrate YARN into scheduler
- #643: metrics tables are now sortable
- #639, fix parallelism is 0 on GlobalGrouping
- #636, fix FieldsGroupingPartitioner
- #629 - query backend for actual websocket address
- #634: create websocket can be failed when url is undefined
- #415, support storm connector
- #613, Show metrics charts in application's detail page
- #546 add a rest api to query WebSocket url
- #624, add description field to TaskDescription
- #434 add api/v1.0 prefix for all rest services
- #618 fix no data returned when call Metrics rest
- #607, bump up kafka version to 0.8.2.1
- #615, Incorrect Dag edge width
- #608 increase the maximum frame size of Akka
- #611: dag looks more elegant
Change List:
- #591, (1) added metrics to application detail page (2) periodically update page contents without pressing refresh (3) replaced angular-dashboard-framework with bootstrap + angular
- #600, Config API should return all config under section "gearpump"
- #597, by default, app wil run for ever except you kill it explicitly.
- #595, use smaller metrics interval
- #417, update READ.ME and refactor DistributedShell
- #388, catch netty channel close exception and warn
- #589, change applicationData's timestamp to 24-h format
- #417, deploy a service across the cluster
- #568, enable history metrics service
- #584, KafkaOffsetManager should only stores one offset per timestamp
- #562 fix AppMaster and Executor restart infinitely
- #576, Add processor level for REST returned DAG, so that the UI can render the DAG correctly.
- #568, [REST] add history metrics service in backend
- #571 AppMaster failed to recover
- #569, dag cannot be rendered
- #566, a) add rest service to shutdown an application b) add a rest service to provide app configuration
- #564, REST should return more data for Application information
- #561, add more task metrics
- #558, [REST]Add missing "name" field for metric Histogram
- #556, Use "Processor" to replace taskGroup in source code
- #555, DAG data returned should contains processor Id
- #544, (1) dashboard will now request appmaster details only once 2) visdag is no longer a widget of adf (3) reduced animation of visdag
- #526, Kafka tests are failing in master branch
- #467, Expose codahale metrics by rest service
- #526, temporary disabled kafka examples to make build pass
- #513, Several remain UI issues in new dashboard, like version tag in dashboard
- #516, Travis failed to deploy the binary to github release
- #510, add log directory for UI server
- #485, retain inactive application history in Master
- #504, 1) AppMaster return more detailed application runtime information. 2) fix a worker bug when returning executors which run on worker.
- #422, [UI] show the number of executors (2) changed layout of application page
- #417, add a experiment module to distribute a zip file to different machines
- #502, fix KafkaStorage loading data for async kafka consumer
- #422, [UI] (1) added home directory in UI (2) removed millis from duration 3) updated dag control height
- #476, fix worker and master log path format in rest service
- #498, worker rest data is not updated in UI
- #397, move distributed shell from experiments to examples folder
- #493 add implicit sender so that the Task can send itself messages[work around]
- #427, use new kafka producer in kafka 0.8.2
- #422, added cluster page to show master and worker information
- #489, make the worker rest information easier to parse
- #202, Add a default serializer for all kinds of messages
- #477, REST Workers should return more information
- #456, uses webjars to serve visjs 3.10.0
- #483, upgrade visdag to 3.10 because widget does not repaint correctly when a node is moved
- #473, Use webjars with spray support instead of bower to serve static resources
- #477, return more data for rest call workers/
- #474, fix rest service UT test fail.
- #479, publish test jar as artifacts
- #419, Reorder the application log by master startup timestamp
- #456, Use visdag to render the graph DAG
- #464, Travis bower integration
- #394, fix kafka producer hang issue
- #468, For test code, the ClockService will throw exception when the DAG is not defined
- #465, fix appname prefix bug
- #461, AppMasterSpec and WorkServiceSpec UT are failing
- #270, Create a force-direct dag
- #453, Add rest service to serve master info
- #423, refactor task by separating TaskActor and Task
- #422, add worker rest api
- #449: avoid load external resource by removing all CDN external links
- #397 refactor distributed shell by using new Cluster API
- #441, ui-portal is failed to build because of spray version conflict
- #430 use application name as unique identifier for an application
- #440, moved dashboard code from conf to service/dashboard
- #402, refactor task manager
- #280, Add websockets to REST api
- #269, Define UI Dashboard structure
- #406, custom executor jvm config in gear.conf is not effective.
- #394, fix ActorSystemBooter process not shut down after application finishes
- #412, add version.sbt
- #410 Add sbt-eclipse plugin and wiki for how to build gearpump on eclipse
- #408, handle Ctrl+C(sigint) gracefully.
- #396, bump up kafka version to 0.8.2.0
- #398, Expose more metrics info
- #389, fix kafka fetch thread bug
- #386 UT fail due to unclosed sequence file
- #370, refactor streaming appmaster
- #358, use uPickle for REST Service
- #372, fix DAG subgraph build and more Graph UT
- #377, fix KafkaSource.pull performance due to the List.append is not having good performance
- #378, construct the serializer explicitly, instead of implicitly via the kryo serializer akka extension
- #380, set the context class loader as the URL class loader for ActorSystem.
- #333, KafkaUtilSpec causes out of memory on travis
- #335, #359, Enable auto-deployment to sonatype
- #299, Some UT may fail randomly, most because of the expectMsg time out
- #338, fix kafka leader not available exception
- #349, scoverage dependencies get into snapshot build binaries.
- #352, add RollingCountSpec
- #356, User's application.conf can not be loaded due to empty config.
- #373, add more restrict checks for warning, deprecation and feature
- #327 fix 0.2.1 build error
- #308 add another dag example project
- #330 Allow user to output the metrics to log file besides graphite
- #244, add more UT,
- #250, refactor kafka implementations
- #301, fix UserConfig ClassNotFoundException
- #306, ClassNotFound for customized Application type
- #299 fix SeqFileStreamProcessorSpec
- #320 fix dead lock in StreamTestUtil.createEchoTaskActor
- #317, allow user to customize the akka system config for appmaster
- #244, add KafkaUtilSpec and kafka integration test
- Published a tech paper about the design on https://typesafe.com/blog/gearpump-real-time-streaming-engine-using-akka
- UT coverage rate to 70%
- Add support for general replay-able data source to support at least once delivery.
- More robust streaming. Be resilient to message loss, message duplication, and zombie processes.
- Refactor Kafka data source for at least once delivery.
- Support general applications besides streaming, add an experimental distrubtedshell application under experiments/.
- Re-defined the AppMaster, and Task interface, It is much easier to write an application now.
- Support submitting and distributing large applications jars.
- Add CI tool, nightly build, code coverage, and defined a formal commit guideline.
- #274 AppMaster cannot connect to worker if there are multiple interface on one machine
- #272 Too many dead log messages in the log
- #266 Kafka grouper conf is incorrect
- #259 fix stream replay api and impl
- #245 jacoco conflict with codecov
- #244 Add more unit test for better test coverage
- #242 Add application submitter username so that we can separate the logs for different user
- #239 REST AppMasterService test failed
- #237 Add more information in log line
- #233 Add TIMEOUT for interactive messages between two parties
- #230 Executor fails to connect with AppMaster when in a multiple NIC environment
- #229 Add a default cluster config under conf/
- #226 return error message to user if not enough slots
- #225 Gearpump logging application logs to different streams(and to persistent storage)
- #222 TimeStampFilter implementation is wrong
- #221 AppMastersDataRequest, AppMasterDataRequest and AppMasterDataDetailRequest should return information about failed appmaster attempts
- #217 Write a small custom AppMaster example, which can run distributed shell.
- #216 Support large application jars
- #215 Improve the API and configs so it more easy to write and submit a new application.
- #213 Add documents about how we do benchmark
- #212 when network partition happen there maybe zombie tasks still sending messages for a while
- #208 Support long-running applications
- #196 When AppMasterStarter fails to load a class, the whole Gearpump cluster crash
- #191 Add docs for all examples
- #184 When packing example to a uber jar, should not include core and streaming jars
- #176 Fix NPE for REST /appmaster/0?detail=true when no appmasters have launched
- #169 Convert REST CustomSerializers to extend DefaultJsonProtocol
- #162 version conflicts between dependencies of sub projects bug
- #148 Revert code to be Java 6 compatible
- #145 Add instructions document for Gearpump commit process
- #138 Update ReadMe.md project description
- #135 Add travis build|passing status icon to README.md
- #130 AppMaster, TaskActor class references should not be explicitly referenced in SubmitApplication and TaskDescription - messages
- #127 Document how to run coverage reports, fix sigar exception seen during tests.
- #117 Fix style and code warnings, use slf4j where appropriate
- #116 Add unit tests to gearpump-core
- #115 Setup travis CI and codecov.io for Gearpump
- #112 Break up examples into separate projects as prerequisite to #100
- #110 Netty Java code cleanup
- #108 Spout and Bolt classes should be renamed
- #106 Add unit tests for REST api
- #104 ActorInitializationException while running wordCount in local mode
- #103 Build error, unable to resolve akka-data-replication