QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation

QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation

ID:40090510

大小:6.46 MB

页数:90页

时间:2019-07-20

QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation_第1页
QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation_第2页
QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation_第3页
QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation_第4页
QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation_第5页
资源描述:

《QConNY2012-NathanMarz-StormDistributedandFaultTolerantRealtimeComputation》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、StormDistributedandfault-tolerantrealtimecomputationNathanMarzTwitterBasicinfo•OpensourcedSeptember19th•Implementationis15,000linesofcode•Usedbyover25companies•>3400watchersonGithub(mostwatchedJVMproject)•Veryactivemailinglist•>4000messages•>1000membersBeforeStormQueuesWorkersExample(simplified

2、)ExampleWorkersschemifytweetsandappendtoHadoopExampleWorkersupdatestatisticsonURLsbyincrementingcountersinCassandraExampleUsemod/hashingtomakesuresameURLalwaysgoestosameworkerScalingDeployReconfigure/redeployProblems•Scalingispainful•Poorfault-tolerance•CodingistediousWhatwewant•Guaranteeddatap

3、rocessing•Horizontalscalability•Fault-tolerance•Nointermediatemessagebrokers!•Higherlevelabstractionthanmessagepassing•“Justworks”StormGuaranteeddataprocessingHorizontalscalabilityFault-toleranceNointermediatemessagebrokers!Higherlevelabstractionthanmessagepassing“Justworks”UsecasesStreamDistr

4、ibutedContinuousprocessingRPCcomputationStormClusterStormClusterMasternode(similartoHadoopJobTracker)StormClusterUsedforclustercoordinationStormClusterRunworkerprocessesStartingatopologyKillingatopologyConcepts•Streams•Spouts•Bolts•TopologiesStreamsTupleTupleTupleTupleTupleTupleTupleUnboundeds

5、equenceoftuplesSpoutsSourceofstreamsSpoutexamples•ReadfromKestrelqueue•ReadfromTwitterstreamingAPIBoltsProcessesinputstreamsandproducesnewstreamsBolts•Functions•Filters•Aggregation•Joins•TalktodatabasesTopologyNetworkofspoutsandboltsTasksSpoutsandboltsexecuteasmanytasksacrosstheclusterTaskexec

6、utionTasksarespreadacrosstheclusterTaskexecutionTasksarespreadacrosstheclusterStreamgroupingWhenatupleisemitted,whichtaskdoesitgoto?Streamgrouping•Shufflegrouping:pickarandomtask•Fieldsgrouping:modhashingonasubsetoftuplefields•Allgrouping:sendtoalltasks•Globalgrouping:picktaskwithlowestidTopolog

7、yshuffle[“id1”,“id2”]shuffle[“url”]shuffleallStreamingwordcountTopologyBuilderisusedtoconstructtopologiesinJavaStreamingwordcountDefineaspoutinthetopologywithparallelismof5tasksStreamingwordcountSplitsentencesintowordswithparallelismof8task

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。
相关文章
更多
相关标签