积分充值
 首页
前端开发
AngularDartElectronFlutterHTML/CSSJavaScriptReactSvelteTypeScriptVue.js构建工具
后端开发
.NetC#C++C语言DenoffmpegGoIdrisJavaJuliaKotlinLeanMakefilenimNode.jsPascalPHPPythonRISC-VRubyRustSwiftUML其它语言区块链开发测试微服务敏捷开发架构设计汇编语言
数据库
Apache DorisApache HBaseCassandraClickHouseFirebirdGreenplumMongoDBMySQLPieCloudDBPostgreSQLRedisSQLSQLiteTiDBVitess数据库中间件数据库工具数据库设计
系统运维
AndroidDevOpshttpdJenkinsLinuxPrometheusTraefikZabbix存储网络与安全
云计算&大数据
Apache APISIXApache FlinkApache KarafApache KyuubiApache OzonedaprDockerHadoopHarborIstioKubernetesOpenShiftPandasrancherRocketMQServerlessService MeshVirtualBoxVMWare云原生CNCF机器学习边缘计算
综合其他
BlenderGIMPKiCadKritaWeblate产品与服务人工智能亿图数据可视化版本控制笔试面试
文库资料
前端
AngularAnt DesignBabelBootstrapChart.jsCSS3EchartsElectronHighchartsHTML/CSSHTML5JavaScriptJerryScriptJestReactSassTypeScriptVue前端工具小程序
后端
.NETApacheC/C++C#CMakeCrystalDartDenoDjangoDubboErlangFastifyFlaskGinGoGoFrameGuzzleIrisJavaJuliaLispLLVMLuaMatplotlibMicronautnimNode.jsPerlPHPPythonQtRPCRubyRustR语言ScalaShellVlangwasmYewZephirZig算法
移动端
AndroidAPP工具FlutterFramework7HarmonyHippyIoniciOSkotlinNativeObject-CPWAReactSwiftuni-appWeex
数据库
ApacheArangoDBCassandraClickHouseCouchDBCrateDBDB2DocumentDBDorisDragonflyDBEdgeDBetcdFirebirdGaussDBGraphGreenPlumHStreamDBHugeGraphimmudbIndexedDBInfluxDBIoTDBKey-ValueKitDBLevelDBM3DBMatrixOneMilvusMongoDBMySQLNavicatNebulaNewSQLNoSQLOceanBaseOpenTSDBOracleOrientDBPostgreSQLPrestoDBQuestDBRedisRocksDBSequoiaDBServerSkytableSQLSQLiteTiDBTiKVTimescaleDBYugabyteDB关系型数据库数据库数据库ORM数据库中间件数据库工具时序数据库
云计算&大数据
ActiveMQAerakiAgentAlluxioAntreaApacheApache APISIXAPISIXBFEBitBookKeeperChaosChoerodonCiliumCloudStackConsulDaprDataEaseDC/OSDockerDrillDruidElasticJobElasticSearchEnvoyErdaFlinkFluentGrafanaHadoopHarborHelmHudiInLongKafkaKnativeKongKubeCubeKubeEdgeKubeflowKubeOperatorKubernetesKubeSphereKubeVelaKumaKylinLibcloudLinkerdLonghornMeiliSearchMeshNacosNATSOKDOpenOpenEBSOpenKruiseOpenPitrixOpenSearchOpenStackOpenTracingOzonePaddlePaddlePolicyPulsarPyTorchRainbondRancherRediSearchScikit-learnServerlessShardingSphereShenYuSparkStormSupersetXuperChainZadig云原生CNCF人工智能区块链数据挖掘机器学习深度学习算法工程边缘计算
UI&美工&设计
BlenderKritaSketchUI设计
网络&系统&运维
AnsibleApacheAWKCeleryCephCI/CDCurveDevOpsGoCDHAProxyIstioJenkinsJumpServerLinuxMacNginxOpenRestyPrometheusServertraefikTrafficUnixWindowsZabbixZipkin安全防护系统内核网络运维监控
综合其它
文章资讯
 上传文档  发布文章  登录账户
IT文库
  • 综合
  • 文档
  • 文章

无数据

分类

全部云计算&大数据(4)Apache Flink(4)

语言

全部英语(4)

格式

全部PDF文档 PDF(4)
 
本次搜索耗时 0.013 秒,为您找到相关结果约 4 个.
  • 全部
  • 云计算&大数据
  • Apache Flink
  • 全部
  • 英语
  • 全部
  • PDF文档 PDF
  • 默认排序
  • 最新排序
  • 页数排序
  • 大小排序
  • 全部时间
  • 最近一天
  • 最近一周
  • 最近一个月
  • 最近三个月
  • 最近半年
  • 最近一年
  • pdf文档 Flow control and load shedding - CS 591 K1: Data Stream Processing and Analytics Spring 2020

    University 2020 22 ??? Vasiliki Kalavri | Boston University 2020 22 Durably buffer events in a channel or source Adjust processing rate of all operators to that of the slowest part of the pipeline If tasks run on different worker nodes, the buffer can be recycled as soon as it is on the TCP channel. • If there is no buffer on the consumer side, reading from the TCP connection is interrupted virtual channels multiplexed over TCP connections: • In the presence of skew, a single overload channel can cause the slowdown of the entire dataflow… can we do better? 25 ??? Vasiliki Kalavri | Boston
    0 码力 | 43 页 | 2.42 MB | 1 年前
    3
  • pdf文档 Streaming optimizations - CS 591 K1: Data Stream Processing and Analytics Spring 2020

    that • Avoid deadlocks: if split cannot push data because one channel is full and merge cannot receive data because another channel is empty Operator fission Data parallelism, replication A A
    0 码力 | 54 页 | 2.83 MB | 1 年前
    3
  • pdf文档 Exactly-once fault-tolerance in Apache Flink - CS 591 K1: Data Stream Processing and Analytics Spring 2020

    the first time: 1. Records its own state. 2. Marks the channel that the marker came in on as empty. a. Future messages arriving on this channel will no be part of the snapshot. 3. Sends markers to its incoming channels except the one marked as empty. Otherwise: 1. It stops recording on the channel it received the marker from. 21 ??? Vasiliki Kalavri | Boston University 2020 Completing a snapshot local state and ll processes have received markers on all incoming channels and have recorded all channel states. We can then collect the locally recorded states and construct a global snapshot. 22
    0 码力 | 81 页 | 13.18 MB | 1 年前
    3
  • pdf文档 Fault-tolerance demo & reconfiguration - CS 591 K1: Data Stream Processing and Analytics Spring 2020

    computation: load in terms of computation • communication: load in terms of flow size in the input channel of each parallel task • Partitioning function performance • space required to implement routing
    0 码力 | 41 页 | 4.09 MB | 1 年前
    3
共 4 条
  • 1
前往
页
相关搜索词
FlowcontrolandloadsheddingCS591K1DataStreamProcessingAnalyticsSpring2020StreamingoptimizationsExactlyoncefaulttoleranceinApacheFlinkFaultdemoreconfiguration
IT文库
关于我们 文库协议 联系我们 意见反馈 免责声明
本站文档数据由用户上传或本站整理自互联网,不以营利为目的,供所有人免费下载和学习使用。如侵犯您的权益,请联系我们进行删除。
IT文库 ©1024 - 2025 | 站点地图
Powered By MOREDOC AI v3.3.0-beta.70
  • 关注我们的公众号【刻舟求荐】,给您不一样的精彩
    关注我们的公众号【刻舟求荐】,给您不一样的精彩