| <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> |
| <!-- |
| | Generated by Apache Maven Doxia at 2021-06-15 |
| | Rendered using Apache Maven Stylus Skin 1.5 |
| --> |
| <html xmlns="http://www.w3.org/1999/xhtml"> |
| <head> |
| <title>Apache Hadoop 3.3.1 – Consistent Reads from HDFS Observer NameNode</title> |
| <style type="text/css" media="all"> |
| @import url("./css/maven-base.css"); |
| @import url("./css/maven-theme.css"); |
| @import url("./css/site.css"); |
| </style> |
| <link rel="stylesheet" href="./css/print.css" type="text/css" media="print" /> |
| <meta name="Date-Revision-yyyymmdd" content="20210615" /> |
| <meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /> |
| </head> |
| <body class="composite"> |
| <div id="banner"> |
| <a href="http://hadoop.apache.org/" id="bannerLeft"> |
| <img src="http://hadoop.apache.org/images/hadoop-logo.jpg" alt="" /> |
| </a> |
| <a href="http://www.apache.org/" id="bannerRight"> |
| <img src="http://www.apache.org/images/asf_logo_wide.png" alt="" /> |
| </a> |
| <div class="clear"> |
| <hr/> |
| </div> |
| </div> |
| <div id="breadcrumbs"> |
| |
| <div class="xleft"> |
| <a href="http://www.apache.org/" class="externalLink">Apache</a> |
| > |
| <a href="http://hadoop.apache.org/" class="externalLink">Hadoop</a> |
| > |
| <a href="../index.html">Apache Hadoop Project Dist POM</a> |
| > |
| <a href="index.html">Apache Hadoop 3.3.1</a> |
| > |
| Consistent Reads from HDFS Observer NameNode |
| </div> |
| <div class="xright"> <a href="http://wiki.apache.org/hadoop" class="externalLink">Wiki</a> |
| | |
| <a href="https://gitbox.apache.org/repos/asf/hadoop.git" class="externalLink">git</a> |
| | |
| <a href="http://hadoop.apache.org/" class="externalLink">Apache Hadoop</a> |
| |
| | Last Published: 2021-06-15 |
| | Version: 3.3.1 |
| </div> |
| <div class="clear"> |
| <hr/> |
| </div> |
| </div> |
| <div id="leftColumn"> |
| <div id="navcolumn"> |
| |
| <h5>General</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../index.html">Overview</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/SingleCluster.html">Single Node Setup</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/ClusterSetup.html">Cluster Setup</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/CommandsManual.html">Commands Reference</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/FileSystemShell.html">FileSystem Shell</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/Compatibility.html">Compatibility Specification</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/DownstreamDev.html">Downstream Developer's Guide</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/AdminCompatibilityGuide.html">Admin Compatibility Guide</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/InterfaceClassification.html">Interface Classification</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/filesystem/index.html">FileSystem Specification</a> |
| </li> |
| </ul> |
| <h5>Common</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/CLIMiniCluster.html">CLI Mini Cluster</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/FairCallQueue.html">Fair Call Queue</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/NativeLibraries.html">Native Libraries</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/Superusers.html">Proxy User</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/RackAwareness.html">Rack Awareness</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/SecureMode.html">Secure Mode</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/ServiceLevelAuth.html">Service Level Authorization</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/HttpAuthentication.html">HTTP Authentication</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/CredentialProviderAPI.html">Credential Provider API</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-kms/index.html">Hadoop KMS</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/Tracing.html">Tracing</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/UnixShellGuide.html">Unix Shell Guide</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/registry/index.html">Registry</a> |
| </li> |
| </ul> |
| <h5>HDFS</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsDesign.html">Architecture</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsUserGuide.html">User Guide</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HDFSCommands.html">Commands Reference</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HDFSHighAvailabilityWithQJM.html">NameNode HA With QJM</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HDFSHighAvailabilityWithNFS.html">NameNode HA With NFS</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/ObserverNameNode.html">Observer NameNode</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/Federation.html">Federation</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/ViewFs.html">ViewFs</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/ViewFsOverloadScheme.html">ViewFsOverloadScheme</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsSnapshots.html">Snapshots</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsEditsViewer.html">Edits Viewer</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsImageViewer.html">Image Viewer</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsPermissionsGuide.html">Permissions and HDFS</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsQuotaAdminGuide.html">Quotas and HDFS</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/LibHdfs.html">libhdfs (C API)</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/WebHDFS.html">WebHDFS (REST API)</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-hdfs-httpfs/index.html">HttpFS</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/ShortCircuitLocalReads.html">Short Circuit Local Reads</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/CentralizedCacheManagement.html">Centralized Cache Management</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsNfsGateway.html">NFS Gateway</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsRollingUpgrade.html">Rolling Upgrade</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/ExtendedAttributes.html">Extended Attributes</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/TransparentEncryption.html">Transparent Encryption</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsMultihoming.html">Multihoming</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/ArchivalStorage.html">Storage Policies</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/MemoryStorage.html">Memory Storage Support</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/SLGUserGuide.html">Synthetic Load Generator</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HDFSErasureCoding.html">Erasure Coding</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HDFSDiskbalancer.html">Disk Balancer</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsUpgradeDomain.html">Upgrade Domain</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsDataNodeAdminGuide.html">DataNode Admin</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs-rbf/HDFSRouterFederation.html">Router Federation</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/HdfsProvidedStorage.html">Provided Storage</a> |
| </li> |
| </ul> |
| <h5>MapReduce</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/MapReduceTutorial.html">Tutorial</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/MapredCommands.html">Commands Reference</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/MapReduce_Compatibility_Hadoop1_Hadoop2.html">Compatibility with 1.x</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/EncryptedShuffle.html">Encrypted Shuffle</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/PluggableShuffleAndPluggableSort.html">Pluggable Shuffle/Sort</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/DistributedCacheDeploy.html">Distributed Cache Deploy</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/SharedCacheSupport.html">Support for YARN Shared Cache</a> |
| </li> |
| </ul> |
| <h5>MapReduce REST APIs</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/MapredAppMasterRest.html">MR Application Master</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-hs/HistoryServerRest.html">MR History Server</a> |
| </li> |
| </ul> |
| <h5>YARN</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/YARN.html">Architecture</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/YarnCommands.html">Commands Reference</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/CapacityScheduler.html">Capacity Scheduler</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/FairScheduler.html">Fair Scheduler</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/ResourceManagerRestart.html">ResourceManager Restart</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/ResourceManagerHA.html">ResourceManager HA</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/ResourceModel.html">Resource Model</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/NodeLabel.html">Node Labels</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/NodeAttributes.html">Node Attributes</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/WebApplicationProxy.html">Web Application Proxy</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/TimelineServer.html">Timeline Server</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/TimelineServiceV2.html">Timeline Service V.2</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/WritingYarnApplications.html">Writing YARN Applications</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/YarnApplicationSecurity.html">YARN Application Security</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/NodeManager.html">NodeManager</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/DockerContainers.html">Running Applications in Docker Containers</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/RuncContainers.html">Running Applications in runC Containers</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/NodeManagerCgroups.html">Using CGroups</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/SecureContainer.html">Secure Containers</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/ReservationSystem.html">Reservation System</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/GracefulDecommission.html">Graceful Decommission</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/OpportunisticContainers.html">Opportunistic Containers</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/Federation.html">YARN Federation</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/SharedCache.html">Shared Cache</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/UsingGpus.html">Using GPU</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/UsingFPGA.html">Using FPGA</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/PlacementConstraints.html">Placement Constraints</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/YarnUI2.html">YARN UI2</a> |
| </li> |
| </ul> |
| <h5>YARN REST APIs</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/WebServicesIntro.html">Introduction</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/ResourceManagerRest.html">Resource Manager</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/NodeManagerRest.html">Node Manager</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/TimelineServer.html#Timeline_Server_REST_API_v1">Timeline Server</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/TimelineServiceV2.html#Timeline_Service_v.2_REST_API">Timeline Service V.2</a> |
| </li> |
| </ul> |
| <h5>YARN Service</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/yarn-service/Overview.html">Overview</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/yarn-service/QuickStart.html">QuickStart</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/yarn-service/Concepts.html">Concepts</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/yarn-service/YarnServiceAPI.html">Yarn Service API</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/yarn-service/ServiceDiscovery.html">Service Discovery</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-site/yarn-service/SystemServices.html">System Services</a> |
| </li> |
| </ul> |
| <h5>Hadoop Compatible File Systems</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-aliyun/tools/hadoop-aliyun/index.html">Aliyun OSS</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-aws/tools/hadoop-aws/index.html">Amazon S3</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-azure/index.html">Azure Blob Storage</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-azure-datalake/index.html">Azure Data Lake Storage</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-openstack/index.html">OpenStack Swift</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-cos/cloud-storage/index.html">Tencent COS</a> |
| </li> |
| </ul> |
| <h5>Auth</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-auth/index.html">Overview</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-auth/Examples.html">Examples</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-auth/Configuration.html">Configuration</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-auth/BuildingIt.html">Building</a> |
| </li> |
| </ul> |
| <h5>Tools</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-streaming/HadoopStreaming.html">Hadoop Streaming</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-archives/HadoopArchives.html">Hadoop Archives</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-archive-logs/HadoopArchiveLogs.html">Hadoop Archive Logs</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-distcp/DistCp.html">DistCp</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-gridmix/GridMix.html">GridMix</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-rumen/Rumen.html">Rumen</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-resourceestimator/ResourceEstimator.html">Resource Estimator Service</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-sls/SchedulerLoadSimulator.html">Scheduler Load Simulator</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/Benchmarking.html">Hadoop Benchmarking</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-dynamometer/Dynamometer.html">Dynamometer</a> |
| </li> |
| </ul> |
| <h5>Reference</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/release/">Changelog and Release Notes</a> |
| </li> |
| <li class="none"> |
| <a href="../../api/index.html">Java API docs</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/UnixShellAPI.html">Unix Shell API</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/Metrics.html">Metrics</a> |
| </li> |
| </ul> |
| <h5>Configuration</h5> |
| <ul> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/core-default.xml">core-default.xml</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs/hdfs-default.xml">hdfs-default.xml</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-hdfs-rbf/hdfs-rbf-default.xml">hdfs-rbf-default.xml</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-mapreduce-client/hadoop-mapreduce-client-core/mapred-default.xml">mapred-default.xml</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-yarn/hadoop-yarn-common/yarn-default.xml">yarn-default.xml</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-kms/kms-default.html">kms-default.xml</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-hdfs-httpfs/httpfs-default.html">httpfs-default.xml</a> |
| </li> |
| <li class="none"> |
| <a href="../../hadoop-project-dist/hadoop-common/DeprecatedProperties.html">Deprecated Properties</a> |
| </li> |
| </ul> |
| <a href="http://maven.apache.org/" title="Built by Maven" class="poweredBy"> |
| <img alt="Built by Maven" src="./images/logos/maven-feather.png"/> |
| </a> |
| |
| </div> |
| </div> |
| <div id="bodyColumn"> |
| <div id="contentBox"> |
| <!--- |
| Licensed under the Apache License, Version 2.0 (the "License"); |
| you may not use this file except in compliance with the License. |
| You may obtain a copy of the License at |
| |
| http://www.apache.org/licenses/LICENSE-2.0 |
| |
| Unless required by applicable law or agreed to in writing, software |
| distributed under the License is distributed on an "AS IS" BASIS, |
| WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| See the License for the specific language governing permissions and |
| limitations under the License. See accompanying LICENSE file. |
| --> |
| <h1>Consistent Reads from HDFS Observer NameNode</h1> |
| <ul> |
| <li><a href="#Purpose">Purpose</a></li> |
| <li><a href="#Background">Background</a></li> |
| <li><a href="#Architecture">Architecture</a> |
| <ul> |
| <li><a href="#Maintaining_Client_Consistency">Maintaining Client Consistency</a></li></ul></li> |
| <li><a href="#Deployment">Deployment</a> |
| <ul> |
| <li><a href="#Configurations">Configurations</a></li> |
| <li><a href="#New_administrative_command">New administrative command</a></li> |
| <li><a href="#Deployment_details">Deployment details</a></li> |
| <li><a href="#Client_configuration">Client configuration</a></li></ul></li></ul> |
| |
| <div class="section"> |
| <h2><a name="Purpose"></a>Purpose</h2> |
| <p>This guide provides an overview of the HDFS Observer NameNode feature and how to configure/install it in a typical HA-enabled cluster. For a detailed technical design overview, please check the doc attached to <a class="externalLink" href="https://issues.apache.org/jira/browse/HDFS-12943">HDFS-12943</a>.</p></div> |
| <div class="section"> |
| <h2><a name="Background"></a>Background</h2> |
| <p>In a HA-enabled HDFS cluster (for more information, check <a href="./HDFSHighAvailabilityWithQJM.html">HDFSHighAvailabilityWithQJM</a>), there is a single Active NameNode and one or more Standby NameNode(s). The Active NameNode is responsible for serving all client requests, while Standby NameNode just keep the up-to-date information regarding the namespace, by tailing edit logs from JournalNodes, as well as block location information, by receiving block reports from all the DataNodes. One drawback of this architecture is that the Active NameNode could be a single bottle-neck and be overloaded with client requests, especially in a busy cluster.</p> |
| <p>The Consistent Reads from HDFS Observer NameNode feature addresses the above by introducing a new type of NameNode called <b>Observer NameNode</b>. Similar to Standby NameNode, Observer NameNode keeps itself up-to-date regarding the namespace and block location information. In addition, it also has the ability to serve consistent reads, like Active NameNode. Since read requests are the majority in a typical environment, this can help to load balancing the NameNode traffic and improve overall throughput.</p></div> |
| <div class="section"> |
| <h2><a name="Architecture"></a>Architecture</h2> |
| <p>In the new architecture, a HA cluster could consists of namenodes in 3 different states: active, standby and observer. State transition can happen between active and standby, standby and observer, but not directly between active and observer.</p> |
| <p>To ensure read-after-write consistency within a single client, a state ID, which is implemented using transaction ID within NameNode, is introduced in RPC headers. When a client performs write through Active NameNode, it updates its state ID using the latest transaction ID from the NameNode. When performing a subsequent read, the client passes this state ID to Observer NameNode, which will then check against its own transaction ID, and will ensure its own transaction ID has caught up with the request’s state ID, before serving the read request. This ensures “read your own writes” semantics from a single client. Maintaining consistency between multiple clients in the face of out-of-band communication is discussed in the “Maintaining Client Consistency” section below.</p> |
| <p>Edit log tailing is critical for Observer NameNode as it directly affects the latency between when a transaction is applied in Active NameNode and when it is applied in the Observer NameNode. A new edit log tailing mechanism, named “Edit Tailing Fast-Path”, is introduced to significantly reduce this latency. This is built on top of the existing in-progress edit log tailing feature, with further improvements such as RPC-based tailing instead of HTTP, a in-memory cache on the JournalNode, etc. For more details, please see the design doc attached to HDFS-13150.</p> |
| <p>New client-side proxy providers are also introduced. ObserverReadProxyProvider, which inherits the existing ConfiguredFailoverProxyProvider, should be used to replace the latter to enable reads from Observer NameNode. When submitting a client read request, the proxy provider will first try each Observer NameNode available in the cluster, and only fall back to Active NameNode if all of the former failed. Similarly, ObserverReadProxyProviderWithIPFailover is introduced to replace IPFailoverProxyProvider in a IP failover setup.</p> |
| <div class="section"> |
| <h3><a name="Maintaining_Client_Consistency"></a>Maintaining Client Consistency</h3> |
| <p>As discussed above, a client ‘foo’ will update its state ID upon every request to the Active NameNode, which includes all write operations. Any request directed to an Observer NameNode will wait until the Observer has seen this transaction ID, ensuring that the client is able to read all of its own writes. However, if ‘foo’ sends an out-of-band (i.e., non-HDFS) message to client ‘bar’ telling it that a write has been performed, a subsequent read by ‘bar’ may not see the recent write by ‘foo’. To prevent this inconsistent behavior, a new <tt>msync()</tt>, or “metadata sync”, command has been added. When <tt>msync()</tt> is called on a client, it will update its state ID against the Active NameNode – a very lightweight operation – so that subsequent reads are guaranteed to be consistent up to the point of the <tt>msync()</tt>. Thus as long as ‘bar’ calls <tt>msync()</tt> before performing its read, it is guaranteed to see the write made by ‘foo’.</p> |
| <p>To make use of <tt>msync()</tt>, an application does not necessarily have to make any code changes. Upon startup, a client will automatically call <tt>msync()</tt> before performing any reads against an Observer, so that any writes performed prior to the initialization of the client will be visible. In addition, there is a configurable “auto-msync” mode supported by ObserverReadProxyProvider which will automatically perform an <tt>msync()</tt> at some configurable interval, to prevent a client from ever seeing data that is more stale than a time bound. There is some overhead associated with this, as each refresh requires an RPC to the Active NameNode, so it is disabled by default.</p></div></div> |
| <div class="section"> |
| <h2><a name="Deployment"></a>Deployment</h2> |
| <div class="section"> |
| <h3><a name="Configurations"></a>Configurations</h3> |
| <p>To enable consistent reads from Observer NameNode, you’ll need to add a few configurations to your <b>hdfs-site.xml</b>:</p> |
| <ul> |
| |
| <li><b>dfs.namenode.state.context.enabled</b> - to enable NameNode to maintain and update server state and id.</li> |
| </ul> |
| <p>This will lead to NameNode creating alignment context instance, which keeps track of current server state id. Server state id will be carried back to client. It is disabled by default to optimize performance of Observer read cases. But this is <b>required to be turned on</b> for the Observer NameNode feature.</p> |
| |
| <div> |
| <div> |
| <pre class="source"> <property> |
| <name>dfs.namenode.state.context.enabled</name> |
| <value>true</value> |
| </property> |
| </pre></div></div> |
| |
| <ul> |
| |
| <li><b>dfs.ha.tail-edits.in-progress</b> - to enable fast tailing on in-progress edit logs.</li> |
| </ul> |
| <p>This enables fast edit log tailing through in-progress edit logs and also other mechanisms such as RPC-based edit log fetching, in-memory cache in JournalNodes, and so on. It is disabled by default, but is <b>required to be turned on</b> for the Observer NameNode feature.</p> |
| |
| <div> |
| <div> |
| <pre class="source"> <property> |
| <name>dfs.ha.tail-edits.in-progress</name> |
| <value>true</value> |
| </property> |
| </pre></div></div> |
| |
| <ul> |
| |
| <li><b>dfs.ha.tail-edits.period</b> - how often Standby/Observer NameNodes should fetch edits from JournalNodes.</li> |
| </ul> |
| <p>This determines the staleness of Observer NameNode w.r.t the Active. If too large, RPC time will increase as client requests will wait longer in the RPC queue before Observer tails edit logs and catches up the latest state of Active. The default value is 1min. It is <b>highly recommend</b> to configure this to a much lower value. It is also recommended to configure backoff to be enabled when using low values; please see below.</p> |
| |
| <div> |
| <div> |
| <pre class="source"> <property> |
| <name>dfs.ha.tail-edits.period</name> |
| <value>0ms</value> |
| </property> |
| </pre></div></div> |
| |
| <ul> |
| |
| <li><b>dfs.ha.tail-edits.period.backoff-max</b> - whether the Standby/Observer NameNodes should perform backoff when tailing edits.</li> |
| </ul> |
| <p>This determines the behavior of a Standby/Observer when it attempts to tail edits from the JournalNodes and finds no edits available. This is a common situation when the edit tailing period is very low, but the cluster is not heavily loaded. Without this configuration, such a situation will cause high utilization on the Standby/Observer as it constantly attempts to read edits even though there are none available. With this configuration enabled, exponential backoff will be performed when an edit tail attempt returns 0 edits. This configuration specifies the maximum time to wait between edit tailing attempts.</p> |
| |
| <div> |
| <div> |
| <pre class="source"> <property> |
| <name>dfs.ha.tail-edits.period.backoff-max</name> |
| <value>10s</value> |
| </property> |
| </pre></div></div> |
| |
| <ul> |
| |
| <li><b>dfs.journalnode.edit-cache-size.bytes</b> - the in-memory cache size, in bytes, on the JournalNodes.</li> |
| </ul> |
| <p>This is the size, in bytes, of the in-memory cache for storing edits on the JournalNode side. The cache is used for serving edits via RPC-based tailing. This is only effective when dfs.ha.tail-edits.in-progress is turned on.</p> |
| |
| <div> |
| <div> |
| <pre class="source"> <property> |
| <name>dfs.journalnode.edit-cache-size.bytes</name> |
| <value>1048576</value> |
| </property> |
| </pre></div></div> |
| |
| <ul> |
| |
| <li><b>dfs.namenode.accesstime.precision</b> – whether to enable access time for HDFS file.</li> |
| </ul> |
| <p>It is <b>highly recommended</b> to disable this configuration. If enabled, this will turn a <tt>getBlockLocations</tt> call into a write call, as it needs to hold write lock to update the time for the opened file. Therefore, the request will fail on all Observer NameNodes and fall back to the active eventually. As result, RPC performance will degrade.</p> |
| |
| <div> |
| <div> |
| <pre class="source"> <property> |
| <name>dfs.namenode.accesstime.precision</name> |
| <value>0</value> |
| </property> |
| </pre></div></div> |
| </div> |
| <div class="section"> |
| <h3><a name="New_administrative_command"></a>New administrative command</h3> |
| <p>A new HA admin command is introduced to transition a Standby NameNode into observer state:</p> |
| |
| <div> |
| <div> |
| <pre class="source">haadmin -transitionToObserver |
| </pre></div></div> |
| |
| <p>Note this can only be executed on Standby NameNode. Exception will be thrown when invoking this on Active NameNode.</p> |
| <p>Similarly, existing <b>transitionToStandby</b> can also be run on an Observer NameNode, which transition it to the standby state.</p> |
| <p><b>NOTE</b>: the feature for Observer NameNode to participate in failover is not implemented yet. Therefore, as described in the next section, you should only use <b>transitionToObserver</b> to bring up an observer. ZKFC could be turned on the Observer NameNode, but it doesn’t do anything when the NameNode is in Observer state. ZKFC will participate in the election of Active after the NameNode is transitioned to standby state.</p></div> |
| <div class="section"> |
| <h3><a name="Deployment_details"></a>Deployment details</h3> |
| <p>To enable observer support, first you’ll need a HA-enabled HDFS cluster with more than 2 namenodes. Then, you need to transition Standby NameNode(s) into the observer state. An minimum setup would be running 3 namenodes in the cluster, one active, one standby and one observer. For large HDFS clusters we recommend running two or more Observers depending on the intensity of read requests and HA requirements.</p> |
| <p>Note that currently Observer NameNode doesn’t integrate fully when automatic failover is enabled. If the <b>dfs.ha.automatic-failover.enabled</b> is turned on, the only benefit for running ZKFC on Observer NameNode is that it will automatically join election of Active after you transition the NameNode to Standby. If this is not desired, you can disable ZKFC on the Observer NameNode. In addition to that, you’ll also need to add <b>forcemanual</b> flag to the <b>transitionToObserver</b> command:</p> |
| |
| <div> |
| <div> |
| <pre class="source">haadmin -transitionToObserver -forcemanual |
| </pre></div></div> |
| |
| <p>In future, this restriction will be lifted.</p></div> |
| <div class="section"> |
| <h3><a name="Client_configuration"></a>Client configuration</h3> |
| <p>Clients who wish to use Observer NameNode for read accesses can specify the ObserverReadProxyProvider class for proxy provider implementation, in the client-side <b>hdfs-site.xml</b> configuration file:</p> |
| |
| <div> |
| <div> |
| <pre class="source"><property> |
| <name>dfs.client.failover.proxy.provider.<nameservice></name> |
| <value>org.apache.hadoop.hdfs.server.namenode.ha.ObserverReadProxyProvider</value> |
| </property> |
| </pre></div></div> |
| |
| <p>Clients who do not wish to use Observer NameNode can still use the existing ConfiguredFailoverProxyProvider and should not see any behavior change.</p> |
| <p>Clients who wish to make use of the “auto-msync” functionality should adjust the configuration below. This will specify some time period after which, if the client’s state ID has not been updated from the Active NameNode, an <tt>msync()</tt> will automatically be performed. If this is specified as 0, an <tt>msync()</tt> will be performed before <i>every</i> read operation. If this is a positive time duration, an <tt>msync()</tt> will be performed every time a read operation is requested and the Active has not been contacted for longer than that period. If this is negative (the default), no automatic <tt>msync()</tt> will be performed.</p> |
| |
| <div> |
| <div> |
| <pre class="source"><property> |
| <name>dfs.client.failover.observer.auto-msync-period.<nameservice></name> |
| <value>500ms</value> |
| </property> |
| </pre></div></div></div></div> |
| </div> |
| </div> |
| <div class="clear"> |
| <hr/> |
| </div> |
| <div id="footer"> |
| <div class="xright"> |
| © 2008-2021 |
| Apache Software Foundation |
| |
| - <a href="http://maven.apache.org/privacy-policy.html">Privacy Policy</a>. |
| Apache Maven, Maven, Apache, the Apache feather logo, and the Apache Maven project logos are trademarks of The Apache Software Foundation. |
| </div> |
| <div class="clear"> |
| <hr/> |
| </div> |
| </div> |
| </body> |
| </html> |
