flink-conf.yaml
Flink 配置文件 对于管理员来说,差不多经常调整的就只有 conf 下的flink-conf.yaml : 经过初步的调整,大约有以下模块的参数(未优化)
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you 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.
==============================================================================
Jobmanager && TaskManager 的基本参数
==============================================================================
jobmanager.rpc.address: flink-master
jobmanager.rpc.port: 6123
jobmanager.heap.size: 32768m
jobstore.cache-size: 52428800
taskmanager.heap.size: 92160m
taskmanager.numberOfTaskSlots: 30
parallelism.default: 3
io.tmp.dirs: /data1/data/flink/local
taskmanager.network.numberOfBuffers: 4096
taskmanager.memory.fraction: 0.7
taskmanager.memory.preallocate: false
taskmanager.debug.memory.startLogThread: 8
taskmanager.debug.memory.logIntervalMs: 3600000
blob.fetch.retries: 50
blob.fetch.num-concurrent: 200
blob.fetch.backlog: 500
task.cancellation-interval: 30000
taskmanager.exit-on-fatal-akka-error: false
taskkmanager.log.path:
==============================================================================
High Availability && and checkpointing
=============================================================================
high-availability: ZOOKEEPER
high-availability.cluster-id: /default_flink_dc
high-availability.zookeeper.quorum: VECS01455:2181,VECS01456:2181,VECS01457:2181,VECS01458:2181,VECS01459:2181
high-availability.storageDir: hdfs://flashHadoop/flink/recovery
high-availability.zookeeper.path.root: /flink
high-availability.zookeeper.path.checkpoint-counter: /checkpoint-counter
high-availability.zookeeper.path.checkpoints: /checkpoints
high-availability.zookeeper.path.jobgraphs: /jobgraphs
high-availability.zookeeper.path.latch: /leaderlatch
high-availability.zookeeper.path.leader: /leader
high-availability.zookeeper.path.mesos-workers: /mesos-workers
high-availability.zookeeper.path.running-registry: /running_job_registry/
high-availability.zookeeper.client.connection-timeout: 18000
high-availability.zookeeper.client.max-retry-attempts: 10
high-availability.zookeeper.client.retry-wait: 5000
high-availability.zookeeper.client.session-timeout: 60000
==============================================================================
HDFS and Fault tolerance and checkpointing
==============================================================================
fs.default-scheme: hdfs://flashHadoop/flink/data
fs.output.always-create-directory: false
fs.overwrite-files: false
state.backend: filesystem
state.backend.async: true
state.backend.fs.memory-threshold: 4096
state.backend.incremental: false
state.backend.local-recovery: false
state.checkpoints.dir: hdfs://flashHadoop/flink/checkpoints_data
state.checkpoints.num-retained: 1
state.savepoints.dir: hdfs://flashHadoop/flink/savepoints_data
==============================================================================
Web Frontend
==============================================================================
web.address: 0.0.0.0
web.access-control-allow-origin: "*"
web.backpressure.cleanup-interval: 600000
web.backpressure.delay-between-samples: 50
web.backpressure.num-samples: 100
web.backpressure.refresh-interval: 60000
web.checkpoints.history: 20
web.history: 20
web.log.path: /data1/data/flink/web/log
web.refresh-interval: 3000
web.submit.enable: true
web.timeout: 10000
web.tmpdir: /data1/data/flink/web/tmp
==============================================================================
Flink Cluster Security Configuration
==============================================================================
test is not set Security
==============================================================================
HistoryServer Config
==============================================================================
historyserver.web.address: 0.0.0.0
historyserver.web.port: 8082
historyserver.archive.fs.dir: hdfs://flashHadoop/flink/completed_jobs/
historyserver.archive.fs.refresh-interval: 2000
historyserver.web.tmpdir: /data1/data/flink/tmp
==============================================================================
==============================================================================
Environment 角色环境 申明
==============================================================================
env.java.opts: -XX:+UseG1GC -XX:+UnlockDiagnosticVMOptions -XX:+UnlockExperimentalVMOptions -XX:+G1SummarizeConcMark -XX:MaxGCPauseMillis=100 -XX:-ResizePLAB -XX:+ParallelRefProcEnabled -XX:ParallelGCThreads=24 -XX:ConcGCThreads=16 -XX:G1HeapWastePercent=3 -XX:InitiatingHeapOccupancyPercent=35 -XX:G1MixedGCLiveThresholdPercent=85
env.java.opts.historyserver: -Xmx16g -Xms16g -Xmn4g -XX:+UseG1GC -XX:+UnlockDiagnosticVMOptions -XX:+UnlockExperimentalVMOptions -XX:+G1SummarizeConcMark -XX:MaxGCPauseMillis=100 -XX:-ResizePLAB -XX:+ParallelRefProcEnabled -XX:+AlwaysPreTouch -XX:ParallelGCThreads=24 -XX:ConcGCThreads=16 -XX:G1HeapWastePercent=3 -XX:InitiatingHeapOccupancyPercent=35 -XX:G1MixedGCLiveThresholdPercent=85 -verbose:gc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -Xloggc:/data1/data/flink/gc/gc-historyserver-hostname.log
env.java.opts.jobmanager: -XX:+UseG1GC -XX:+UnlockDiagnosticVMOptions -XX:+UnlockExperimentalVMOptions -XX:+G1SummarizeConcMark -XX:MaxGCPauseMillis=100 -XX:-ResizePLAB -XX:+ParallelRefProcEnabled -XX:ParallelGCThreads=24 -XX:ConcGCThreads=16 -XX:G1HeapWastePercent=3 -XX:InitiatingHeapOccupancyPercent=35 -XX:G1MixedGCLiveThresholdPercent=85 -verbose:gc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -Xloggc:/data1/data/flink/gc/gc-jobmanager-hostname.log
env.java.opts.taskmanager: -XX:+UseG1GC -XX:+UnlockDiagnosticVMOptions -XX:+UnlockExperimentalVMOptions -XX:+G1SummarizeConcMark -XX:MaxGCPauseMillis=100 -XX:-ResizePLAB -XX:+ParallelRefProcEnabled -XX:ParallelGCThreads=24 -XX:ConcGCThreads=16 -XX:G1HeapWastePercent=3 -XX:InitiatingHeapOccupancyPercent=35 -XX:G1MixedGCLiveThresholdPercent=85 -verbose:gc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -Xloggc:/data1/data/flink/gc/gc-taskmanager-hostname.log
env.log.dir: /data1/data/flink/log
env.log.max: 1
flink-conf.yaml的更多相关文章
- flink部署操作-flink standalone集群安装部署
flink集群安装部署 standalone集群模式 必须依赖 必须的软件 JAVA_HOME配置 flink安装 配置flink 启动flink 添加Jobmanager/taskmanager 实 ...
- flink Standalone Cluster
Requirements Software Requirements Flink runs on all UNIX-like environments, e.g. Linux, Mac OS X, a ...
- Flink集群部署
部署方式 一般来讲有三种方式: Local Standalone Flink On Yarn/Mesos/K8s… 单机模式 参考上一篇Flink从入门到放弃(入门篇2)-本地环境搭建&构建第 ...
- Flink(二)CentOS7.5搭建Flink1.6.1分布式集群
一. Flink的下载 安装包下载地址:http://flink.apache.org/downloads.html ,选择对应Hadoop的Flink版本下载 [admin@node21 soft ...
- Flink的安装配置
一. Flink的下载 安装包下载地址:http://flink.apache.org/downloads.html ,选择对应Hadoop的Flink版本下载 [admin@node21 soft ...
- 一张图轻松掌握 Flink on YARN 应用启动全流程(上)
Flink 支持 Standalone 独立部署和 YARN.Kubernetes.Mesos 等集群部署模式,其中 YARN 集群部署模式在国内的应用越来越广泛.Flink 社区将推出 Flink ...
- kerberos系列之flink认证配置
大数据安全系列的其它文章 https://www.cnblogs.com/bainianminguo/p/12548076.html-----------安装kerberos https://www. ...
- 分享一个Flink checkpoint失败的问题和解决办法
本文来自: PerfMa技术社区 PerfMa(笨马网络)官网 接触Flink一段时间了,遇到了一些问题,其中有一个checkpoint失败导致作业重启的问题,遇到了很多次,重启之后一般也能恢复正常, ...
- flink集群模式安装配置
一.手动下载安装包 wget http://mirrors.tuna.tsinghua.edu.cn/apache/flink/flink-1.6.1/flink-1.6.1-bin-hadoop27 ...
- Linux使用docker安装flink
配置文件不是原始的 修改过,可以根据自己的配置文件来 jobmanager配置 flink-conf.yaml ############################################ ...
随机推荐
- rocketmq4.x快速入门指南
以下采用的是apache rocketmq 4.2.0版本 相关文档如下 快速体验: http://blog.seoui.com/2018/07/24/rocketmqinstall/ rocketm ...
- ORM初探(一)
Object Relational Mapping(ORM): 对象关系映射(Object Relational Mapping,简称ORM)模式是一种为了解决面向对象与关系数据库存在的互不匹配的现象 ...
- SpringBoot使用qq邮箱发送邮件
最近公司要做一个邮箱注册和重置密码的功能,因为之前就做过,但是不是Springboot项目,所以相对来说还是比较容易的,在这里记录一下. 一.引用Maven依赖 这里使用spring自带的邮件jar包 ...
- Linux使用IDEA配置maven的web项目骨架archetype(模板) 自定义骨架
说明:本文说的骨架就是 archetype,也可以理解为模板,总是就是指你创建项目时的基本配置. 前言:在使用IDEA创建maven的web项目时,一般都是直接使用提供的默认web项目,如图 然而创建 ...
- AI 学习路线
[导读] 本文由知名开源平台,AI技术平台以及领域专家:Datawhale,ApacheCN,AI有道和黄海广博士联合整理贡献,内容涵盖AI入门基础知识.数据分析挖掘.机器学习.深度学习.强化学习.前 ...
- KL散度、JS散度、Wasserstein距离
1. KL散度 KL散度又称为相对熵,信息散度,信息增益.KL散度是是两个概率分布 $P$ 和 $Q$ 之间差别的非对称性的度量. KL散度是用来 度量使用基于 $Q$ 的编码来编码来自 $P$ 的 ...
- html简单的知识
分布式版本控制git pwd查询当前目录 ls ls -la git config --global user.name xxx git config --global user. ...
- Robust PCA via Outlier Pursuit
目录 引 主要结果 定理1 定理2 理论证明 构造Oracle Problem 算法 Xu H, Caramanis C, Sanghavi S, et al. Robust PCA via Outl ...
- Neutron vxlan network--L2 Population
L2 Population 是用来提高 VXLAN 网络 Scalability 的. 通常我们说某个系统的 Scalability 好,其意思是: 当系统的规模变大时,仍然能够高效地工作. L2 ...
- 石家庄地铁系统开发(java web版)(二)
两种方法: 一,自己写数据库,自己写算法实现 二,调用已有软件的API(百度,高德)