Spark资源管理

1、介绍

Spark资源管控分为spark集群自身可支配资源配置和job所用资源配置。

2、spark集群支配资源控制

在spark的conf/spark-env.sh文件中可以指定master和worker的支配资源数。

2.1 Spark集群可支配资源配置

每个worker使用内核数

# 每个worker使用的内核数，默认是所有内核。

export SPARK_WORKER_CORES=6

每个worker所用内存数

# 每个worker使用的内存数，默认是1g内存

export SPARK_WORKER_MEMORY=6g

每个节点可以启动worker实例的个数

#是否可以在一个节点启动几个worker进程，默认1

export SPARK_WORKER_INSTANCES=2

spark守护进程本身占用的内存数

spark守护进程指的是master和worker进程，该进程自身使用内存数也可以进行控制。
```
#master和worker进程本身的内存数 ,默认1g

export SPARK_DAEMON_MEMORY=200m
```

spark/conf/spark-env.sh配置全部内容如下：

#!/usr/bin/env bash

#

# 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.

#

# This file is sourced when running various Spark programs.

# Copy it as spark-env.sh and edit that to configure Spark for your site.

export JAVA_HOME=/soft/jdk

# Options read when launching programs locally with

# ./bin/run-example or ./bin/spark-submit

# - HADOOP_CONF_DIR, to point Spark towards Hadoop configuration files

# export HADOOP_CONF_DIR=/soft/hadoop/etc/hadoop

# - SPARK_LOCAL_IP, to set the IP address Spark binds to on this node

# - SPARK_PUBLIC_DNS, to set the public dns name of the driver program

# - SPARK_CLASSPATH, default classpath entries to append

# Options read by executors and drivers running inside the cluster

# - SPARK_LOCAL_IP, to set the IP address Spark binds to on this node

# - SPARK_PUBLIC_DNS, to set the public DNS name of the driver program

# - SPARK_CLASSPATH, default classpath entries to append

# - SPARK_LOCAL_DIRS, storage directories to use on this node for shuffle and RDD data

# - MESOS_NATIVE_JAVA_LIBRARY, to point to your libmesos.so if you use Mesos

# Options read in YARN client mode

# - HADOOP_CONF_DIR, to point Spark towards Hadoop configuration files

# - SPARK_EXECUTOR_INSTANCES, Number of executors to start (Default: 2)

# - SPARK_EXECUTOR_CORES, Number of cores for the executors (Default: 1).

# - SPARK_EXECUTOR_MEMORY, Memory per Executor (e.g. 1000M, 2G) (Default: 1G)

# - SPARK_DRIVER_MEMORY, Memory for Driver (e.g. 1000M, 2G) (Default: 1G)

# Options for the daemons used in the standalone deploy mode

# - SPARK_MASTER_HOST, to bind the master to a different IP address or hostname

# - SPARK_MASTER_PORT / SPARK_MASTER_WEBUI_PORT, to use non-default ports for the master

#export SPARK_MASTER_PORT=7077

#export SPARK_MASTER_WEBUI_PORT=8080

# - SPARK_MASTER_OPTS, to set config properties only for the master (e.g. "-Dx=y")

# - SPARK_WORKER_CORES, to set the number of cores to use on this machine

export SPARK_WORKER_CORES=6

# - SPARK_WORKER_MEMORY, to set how much total memory workers have to give executors (e.g. 1000m, 2g)

export SPARK_WORKER_MEMORY=6g

# - SPARK_WORKER_PORT / SPARK_WORKER_WEBUI_PORT, to use non-default ports for the worker

# - SPARK_WORKER_INSTANCES, to set the number of worker processes per node

export SPARK_WORKER_INSTANCES=1

# - SPARK_WORKER_DIR, to set the working directory of worker processes

# - SPARK_WORKER_OPTS, to set config properties only for the worker (e.g. "-Dx=y")

# - SPARK_DAEMON_MEMORY, to allocate to the master, worker and history server themselves (default: 1g).

export SPARK_DAEMON_MEMORY=200m

# - SPARK_HISTORY_OPTS, to set config properties only for the history server (e.g. "-Dx=y")

# - SPARK_SHUFFLE_OPTS, to set config properties only for the external shuffle service (e.g. "-Dx=y")

# - SPARK_DAEMON_JAVA_OPTS, to set config properties for all daemons (e.g. "-Dx=y")

# export SPARK_DAEMON_JAVA_OPTS="-Dspark.deploy.recoveryMode=ZOOKEEPER -Dspark.deploy.zookeeper.url=s102:2181,s103:2181,s104:2181 -Dspark.deploy.zookeeper.dir=/spark"

# - SPARK_PUBLIC_DNS, to set the public dns name of the master or workers

# Generic options for the daemons used in the standalone deploy mode

# - SPARK_CONF_DIR      Alternate conf dir. (Default: ${SPARK_HOME}/conf)

# - SPARK_LOG_DIR       Where log files are stored.  (Default: ${SPARK_HOME}/logs)

# - SPARK_PID_DIR       Where the pid file is stored. (Default: /tmp)

# - SPARK_IDENT_STRING  A string representing this instance of spark. (Default: $USER)

# - SPARK_NICENESS      The scheduling priority for daemons. (Default: 0)

# - SPARK_NO_DAEMONIZE  Run the proposed command in the foreground. It will not output a PID file.

2.2 job资源分配设置

spark-submit命令提交job时，可以为job指定使用的资源，包括内存和内核数。但在不同的spark集群模式下，使用的配置命令是不同的。命令使用如下：

$>spark-submit --master spark://s101:7077 --executor-memory 200m --executor-cores 4

设置driver内存数

$>spark-submit --driver-cores 2		#默认是1

standalone和mesos模式下
- 执行器内核总数设置
```
$>spark-submit --total-executor-cores 32
```

standalone和yarn模式

每个执行器内核数设置

# yarn下模式为1，standalone模式下为worker上可有的所有内核数

$>spark-submit --executor-cores 4

yarn-only

只在yarn模式下，使用的资源控制选线：
- driver内核数设置
  
  driver使用的cpu内核数，只在cluster模式下有效，默认1。
```
$>spark-submit --driver-cores 3
```
- 执行器个数设置
  
  启动的执行器个数，默认为2。
```
$>spark-submit --num-executors 3
```

3、资源控制细则

spark资源控制在集群配置时不进行物理资源检查，即可以配置任意的资源值。比如物理内核是16，但是配置成每个worker占用32核。如图所示：

图中箭头指向的部分是worker进程能够支配使用的资源，包括内存和内核数。

Spark job执行时，指定的资源同时受到内存和内核两方面的限制，即任何一个条件不满足，都无法启动executor进程。例如指定每个executor使用3个core，worker可以支配8个core，但是最终该worker只能启动两个executor。