Centos7安装pbs

2023-12-16
1分钟阅读时长

机器配置

主机名 ip
master 192.168.2.1
node01 192.168.2.2
node02 192.168.2.3
node03 192.168.2.4
node04 192.168.2.5
node05 192.168.2.6
node07 192.168.2.7

安装

配置好主机名和SSH互相免密

安装依赖

yum -y install libX11-devel libXt-devel libedit-devel libical-devel ncurses-devel postgresql-devel python-devel python-devel tcl-devel tk-devel swig expat-devel openssl-devel libXext libXft gcc autoconf automake libtool hwloc-devel rpm-build

pbs github地址

Releases · openpbs/openpbs (github.com)

wget https://github.com/openpbs/openpbs/releases/download/v19.1.3/pbspro_19.1.3.centos_7.zip
tar zxvf pbspro_19.1.3.centos_7.zip

设置环境变量 追加到/etc/profile

export PBS_SERVER=master
export PBS_EXEC=/opt/pbs
export PBS_HOME=/var/spool/pbs

master节点 安装 pbspro-server-19.1.3-0.x86_64.rpm

node节点 安装 pbspro-execution-19.1.3-0.x86_64.rpm

master节点和node节点依次启动pbs

systemctl start pbs

如果想让master节点也参与计算,将/etc/pbs.conf 中 PBS_START_MOM的值由0修改为1

配置

master节点其他配置

qmgr -c "set server flatuid=true"
qmgr -c "set server scheduling=true"
qmgr -c "s s job_history_enable=1"

master节点添加节点node01

qmgr -c "create node node01"

查看节点状态

pbsnodes -aSj

image-20241114105950692

master节点创建队列big

qmgr -c "create queue big"

将node7添加到队列big

qmgr -c "set node node7 queue=big"

设置big队列资源限制

qmgr -c "set queue big resources_max.ncpus = 80"
qmgr -c "set queue big resources_max.mem = 503"
qmgr -c "set queue big resources_max.nodect = 1"

启用big队列

qmgr -c "set queue big enabled = true"
qmgr -c "set queue big started = true"

查看big队列参数

qstat -Q -f big

image-20241114110017178

注意 普通用户提交作业前需要配置和计算节点间的SSH互相免密