列出的所有端口都是TCP。

在下表中,每个端口的“ 访问要求”列通常是“内部”或“外部”。在此上下文中,“内部”表示端口仅用于组件之间的通信; “外部”表示该端口可用于内部或外部通信。
Component Service Qualifier Port Access Requirement Configuration Comment

Hadoop HDFS

DataNode

50010

External

dfs.datanode.address

DataNode HTTP server port

DataNode

Secure

1004

External

dfs.datanode.address

DataNode

50075

External

dfs.datanode.http.address

DataNode

50475

External

dfs.datanode.https.address

DataNode

Secure

1006

External

dfs.datanode.http.address

DataNode

50020

External

dfs.datanode.ipc.address

NameNode

8020

External

fs.default.name

or

fs.defaultFS
fs.default.name

is deprecated (but still works)

NameNode

8022

External

dfs.namenode. servicerpc-address

Optional port used by HDFS daemons to avoid sharing the RPC port used by clients (8020). Cloudera recommends using port 8022.

NameNode

50070

External

dfs.http.address

or

dfs.namenode.http-address
dfs.http.address

is deprecated (but still works)

NameNode

Secure

50470

External

dfs.https.address

or

dfs.namenode.https-address
dfs.https.address

is deprecated (but still works)

Secondary NameNode

50090

Internal

dfs.secondary.http.address

or

dfs.namenode. secondary.http-address
dfs.secondary.http.address

is deprecated (but still works)

Secondary NameNode

Secure

50495

Internal

dfs.secondary.https.address

JournalNode

8485

Internal

dfs.namenode.shared.edits.dir

JournalNode

8480

Internal

dfs.journalnode. http-address

JournalNode

8481

Internal

dfs.journalnode. https-address

Failover Controller

8019

Internal

Used for NameNode HA

NFS gateway

2049

External

nfs port (nfs3.server.port)

NFS gateway

4242

External

mountd port (nfs3.mountd.port)

NFS gateway

111

External

portmapper

or

rpcbind

port

  NFS gateway   50079 External
nfs.http.port
CDH 5.4.0 and higher. The NFS gateway daemon uses this port to serve metrics. The port is configurable on versions 5.10 and higher.
  NFS gateway Secure 50579 External
nfs.https.port
CDH 5.4.0 and higher. The NFS gateway daemon uses this port to serve metrics. The port is configurable on versions 5.10 and higher.

HttpFS

14000

External    

HttpFS

14001

External    

Hadoop YARN (MRv2)

ResourceManager

8032

External

yarn. resourcemanager.address

ResourceManager

8030

Internal

yarn. resourcemanager.scheduler.address

ResourceManager

8031

Internal

yarn. resourcemanager.resource-tracker. address

ResourceManager

8033

External

yarn. resourcemanager.admin.address

ResourceManager

8088

External

yarn. resourcemanager.webapp.address

ResourceManager

8090

External

yarn. resourcemanager.webapp.https.address

NodeManager

8040

Internal

yarn. nodemanager.localizer. address

NodeManager

8041

Internal

yarn. nodemanager. address

NodeManager

8042

External

yarn. nodemanager.webapp.address

NodeManager

8044

External

yarn. nodemanager.webapp.https.address

JobHistory Server

10020

Internal

mapreduce. jobhistory.address

JobHistory Server

10033

Internal

mapreduce.jobhistory.admin. address

Shuffle HTTP

13562

Internal

mapreduce.shuffle.port

JobHistory Server

19888

External

mapreduce. jobhistory.webapp.address

JobHistory Server

19890

External

mapreduce. jobhistory.webapp.https.address
 

ApplicationMaster

   

External

 

The ApplicationMaster serves an HTTP service using an ephemeral port that cannot be restricted. This port is never accessed directly from outside the cluster by clients. All requests to the ApplicationMaster web server is routed using the YARN ResourceManager (proxy service). Locking down access to ephemeral port ranges within the cluster's network might restrict your access to the ApplicationMaster UI and its logs, along with the ability to look at running applications.

Flume

Flume Agent

41414

External

   

Hadoop KMS

Key Management Server

16000

External

kms_http_port

CDH 5.2.1 and higher. Applies to both Java KeyStore KMS and Key Trustee KMS.

 

Key Management Server

16001

Localhost

kms_admin_port

CDH 5.2.1 and higher. Applies to both Java KeyStore KMS and Key Trustee KMS.

HBase

Master

60000

External

hbase.master. port

IPC

Master

60010

External

hbase.master. info.port

HTTP

RegionServer

60020

External

hbase. regionserver. port

IPC

RegionServer

60030

External

hbase. regionserver.info.port

HTTP

HQuorumPeer

2181

Internal

hbase. zookeeper. property.clientPort

HBase-managed ZooKeeper mode

HQuorumPeer

2888

Internal

hbase. zookeeper. peerport

HBase-managed ZooKeeper mode

HQuorumPeer

3888

Internal

hbase. zookeeper.leaderport

HBase-managed ZooKeeper mode

REST

Non- Cloudera Manager - managed

8080

External

hbase.rest.port

The default REST port in HBase is 8080. Because this is a commonly used port, Cloudera Manager sets the default to 20550 instead.

REST

Cloudera Manager - managed

20550

External

hbase.rest.port

The default REST port in HBase is 8080. Because this is a commonly used port, Cloudera Manager sets the default to 20550 instead.

REST UI

 

8085

External

 

Thrift Server

Thrift Server

9090

External

Pass -p <port> on CLI

Thrift Server

 

9095

External

 

Avro server

9090

External

Pass --port <port> on CLI

  hbase-solr-indexer Lily Indexer 11060 External    

Hive

Metastore

9083

External

HiveServer2

10000

External

hive. server2. thrift.port

The Beeline command interpreter requires that you specify this port on the command line.

If you use Oracle database, you must manually reserve this port. For more information, see Reserving Ports for HiveServer 2.

HiveServer2 Web User Interface (UI)

10002

External

hive. server2. webui.port

in

hive-site.xml
 

WebHCat Server

50111

External

templeton.port

Hue

Server

8888

External

Kafka

Broker

TCP Port

9092

External/Internal

port

The primary communication port used by producers and consumers; also used for inter-broker communication.

 

Broker

TLS/SSL Port

9093

External/Internal

ssl_port

A secured communication port used by producers and consumers; also used for inter-broker communication.

 

Broker

JMX Port

9393

Internal

jmx_port

Internal use only. Used for administration via JMX.

 

MirrorMaker

JMX Port

9394

Internal

jmx_port

Internal use only. Used to administer the producer and consumer of the MirrorMaker.

 

Broker

HTTP Metric Report Port

24042

Internal

kafka.http.metrics.port

Internal use only. This is the port via which the HTTP metric reporter listens. It is used to retrieve metrics through HTTP instead of JMX.

Kudu

Master

 

7051

External

 

Kudu Master RPC port

 

Master

 

8051

External

 

Kudu Master HTTP server port

 

TabletServer

 

7050

External

 

Kudu TabletServer RPC port

 

TabletServer

 

8050

External

 

Kudu TabletServer HTTP server port

Oozie

Oozie Server

11000

External

OOZIE_HTTP_PORT

in

oozie-env.sh

HTTP

 

Oozie Server

SSL

11443

External

 

HTTPS

Sentry

Sentry Server

8038

External

sentry.service. server.rpc-port
 

Sentry Server

51000

External

sentry.service. web.port

Spark

Default Master RPC port

 

7077

External

   
 

Default Worker RPC port

 

7078

External    
 

Default Master web UI port

 

18080

External

   
 

Default Worker web UI port

 

18081

External    
 

History Server

18088

External

history.port
 
  Shuffle service   7337 Internal    

Sqoop

Metastore

16000

External

sqoop. metastore.server.port

Sqoop 2

Sqoop 2 server

 

8005

Localhost

SQOOP_ADMIN_PORTenvironment variable

 
 

Sqoop 2 server

12000

External

 

Sqoop 2

12001

External

Admin port

ZooKeeper

Server (with CDH 5 or Cloudera Manager 5)

2181

External

clientPort

Client port

Server (with CDH 5 only)

2888

Internal

X in server.N =host:X:Y

Peer

Server (with CDH 5 only)

3888

Internal

X in server.N =host:X:Y

Peer

Server (with CDH 5 and Cloudera Manager 5)

3181

Internal

X in server.N =host:X:Y

Peer

Server (with CDH 5 and Cloudera Manager 5)

4181

Internal

X in server.N =host:X:Y

Peer

ZooKeeper JMX port

9010

Internal

ZooKeeper will also use another randomly selected port for RMI. To allow Cloudera Manager to monitor ZooKeeper, you must do oneof the following:

  • Open up all ports when the connection originates from the Cloudera Manager Server
  • Do the following:
    1. Open a non-ephemeral port (such as 9011) in the firewall.
    2. Install Oracle Java 7u4 JDK or higher.
    3. Add the port configuration to the advanced configuration snippet, for example:
      -Dcom.sun.management.jmxremote.rmi.port=9011
      
    4. Restart ZooKeeper.

1.3.3、CDH 搭建Hadoop在安装之前(端口---CDH组件使用的端口)的更多相关文章

  1. 1.3、CDH 搭建Hadoop在安装之前(端口)

    端口 Cloudera Manager,CDH组件,托管服务和第三方组件使用下表中列出的端口.在部署Cloudera Manager,CDH和托管服务以及第三方组件之前,请确保在每个系统上打开这些端口 ...

  2. 1.5.7、CDH 搭建Hadoop在安装之前(定制安装解决方案---配置单用户模式)

    配置单用户模式 在传统的Cloudera Manager部署中,管理每台主机上的Hadoop进程的Cloudera Manager Agent以root用户身份运行.但是,某些环境会限制对root帐户 ...

  3. 2.6、CDH 搭建Hadoop在安装(安装CDH和其他软件)

    第6步:安装CDH和其他软件 设置Cloudera Manager数据库后,启动Cloudera Manager Server,然后登录Cloudera Manager Admin Console: ...

  4. 2.1、CDH 搭建Hadoop在安装(为Cloudera Manager配置存储库)

    步骤1:为Cloudera Manager配置存储库 使用包管理工具安装Cloudera Manager yum 对于RHEL兼容系统, zypper对于SLES,和 apt-get对于Ubuntu. ...

  5. 1.2、CDH 搭建Hadoop在安装之前(CDH基于包的安装所需的权限)

    CDH基于包的安装所需的权限 以下部分描述了使用Cloudera Manager进行基于软件包的CDH安装的用户权限要求.这些要求是安装和管理包和服务的标准UNIX系统要求. 所需特权 sudo由Cl ...

  6. 1.3.1、CDH 搭建Hadoop在安装之前(端口---Cloudera Manager和Cloudera Navigator使用的端口)

    下图概述了Cloudera Manager,Cloudera Navigator和Cloudera Management Service角色使用的一些端口: Cloudera Manager和Clou ...

  7. 2.2、CDH 搭建Hadoop在安装(安装Java Development Kit)

    第2步:安装Java Development Kit 要安装Oracle JDK,您可以使用Cloudera Manager安装Cloudera提供的版本,也可以直接安装Oracle的其他版本. 继续 ...

  8. 1.4、CDH 搭建Hadoop在安装之前(推荐的群集主机和角色分配)

    推荐的群集主机和角色分配 要点:本主题描述了Cloudera Manager管理的CDH群集的建议角色分配.您为部署选择的实际分配可能会有所不同,具体取决于工作负载的类型和数量,群集中部署的服务,硬件 ...

  9. 2.7、CDH 搭建Hadoop在安装(使用向导设置群集)

    步骤7:使用向导设置群集 完成“ 群集安装”向导后,“ 群集设置”向导将自动启动.以下部分将指导您完成向导的每个页面: 选择服务 分配角色 设置数据库 查看更改 首次运行命令 恭喜! 选择服务 “ 选 ...

  10. 2.5、CDH 搭建Hadoop在安装(设置Cloudera Manager数据库)

    步骤5:设置Cloudera Manager数据库 Cloudera Manager Server包含一个可以为自己创建和配置数据库的脚本.该脚本可以: 创建Cloudera Manager Serv ...

随机推荐

  1. 【Jenkins学习】安装配置和使用(一)

    为了能够频繁地将软件的最新版本,及时.持续地交付给测试团队及质量控制团队,以供评审,所以引入持续集成工具Jenkins,从而实现公司新产品持续集成,自动化部署. 环境准备 ●操作系统:Windows1 ...

  2. android 开发 在一个工具类(或者适配器class)里启动activity

    实现思路: 1.需要给工具类里传入context: 2.使用上下文mContext.startActivity启动activity 例子1: public class SafePlaceRecycle ...

  3. tornado-模版reverse_url

    from tornado.web import url handler=[ url(r"/hello", grab_handler.HelloHandler, name='hell ...

  4. 快速了解CSS3当中的HSLA 颜色值怎么算

    CSS3文档中提到:(HSLA) H是色度,取值在0度~360度之间,0度是红色,120度是绿色,240度是蓝色.360度也是红色. S是饱和度,是色彩的纯度,是一个百分比的值,取值在0%~100%, ...

  5. 【Python爬虫实战】Scrapy框架的安装 搬运工亲测有效

    windows下亲测有效 http://blog.csdn.net/liuweiyuxiang/article/details/68929999这个我们只是正确操作步骤详解的搬运工

  6. 【Python】python3 正则爬取网页输出中文乱码解决

    爬取网页时候print输出的时候有中文输出乱码 例如: \\xe4\\xb8\\xad\\xe5\\x8d\\x8e\\xe4\\xb9\\xa6\\xe5\\xb1\\x80 #爬取https:// ...

  7. pm2操作总结

    PM2是一个node.js的进程管理器,(并且呢在应用程序的生产运行时自带负载均衡的这种操作,很厉害): -->  pm2主要解决的问题是kill node进程时无法正常停止的问题. 主要特征: ...

  8. 【转】Django 图表制作(By Highcharts)

    马克,待不时之需 Django 图表制作(By Highcharts):https://blog.csdn.net/Temanm/article/details/54141759 免费而优秀的图表JS ...

  9. day27-反射

    1.介绍 反射的概念是由Smith在1982年首次提出的,主要是指程序可以访问.检测和修改它本身状态或行为的一种能力(自省).这一概念的提出很快引发了计算机科学领域关于应用反射性的研究.它首先被程序语 ...

  10. mui集成百度ECharts的统计图表以及清空释放图表

    echarts官网地址: http://echarts.baidu.com/index.html 更换主题颜色: // 图表清空------------------- mychart.clear(); ...