Vertica 7.1安装最佳实践（RHEL6.4）

2020年7月12日 21:04:14Linux大全评论261 views阅读模式

一、前期准备工作

1.1各节点IP和主机名
1.2上传脚本并设定环境变量
1.3添加信任
1.4前期准备检查并调整

二、Vertica安装

三、集群性能评估

一、前期准备工作：

1.1各节点IP和主机名

192.168.1.137 DB01
192.168.1.138 DB02
192.168.1.139 DB03
192.168.1.140 DB04

在节点1配置/etc/hosts,添加上面信息。

1.2上传脚本并设定环境变量

在节点1上传两个安装脚本到/usr/local/bin

cluster_copy_all_nodes

#!/bin/bash
SELF=`hostname`
if [ -z "$NODE_LIST" ]; then
  echo
  echo Error: NODE_LIST environment variable must be set in .bash_profile
  exit 1
fi

for i in $NODE_LIST; do
  if [ ! $i = $SELF ]; then
    if [ $1 = "-r" ]; then
      scp -oStrictHostKeyChecking=no -r $2 $i:$3
    else
      scp -oStrictHostKeyChecking=no $1 $i:$2
    fi
  fi
done
wait

cluster_run_all_nodes

#!/bin/bash
if [ -z "$NODE_LIST" ]; then
  echo
  echo Error: NODE_LIST environment variable must be set in .bash_profile
  exit 1
fi

if [[ $1 = '--background' ]]; then
  shift
  for i in $NODE_LIST; do
    ssh -oStrictHostKeyChecking=no -n $i "$@" &
  done
else
  for i in $NODE_LIST; do
    ssh -oStrictHostKeyChecking=no $i "$@"
  done
fi
wait

配置节点1的环境变量

export NODE_LIST='DB01 DB02 DB03 DB04'

1.3添加信任

1 2 3 4 5 6 7 8 ssh - keygen - q - t rsa - N "" - f ~ / .ssh / id_rsa ssh - copy - id - i / root / .ssh / id_rsa.pub root@ 192.168 . 1.137 ssh - copy - id - i / root / .ssh / id_rsa.pub root@ 192.168 . 1.138 ssh - copy - id - i / root / .ssh / id_rsa.pub root@ 192.168 . 1.139 ssh - copy - id - i / root / .ssh / id_rsa.pub root@ 192.168 . 1.140 cluster_run_all_nodes "hostname ; date"

1.4前期准备检查并调整

1.4.1 同步检查系统版本，运行级别，挂载目录结构，网卡信息

1 2 3 4 cluster_run_all_nodes "hostname;cat /etc/RedHat-release" cluster_run_all_nodes "hostname; /sbin/runlevel " cluster_run_all_nodes "hostname; df -h" cluster_run_all_nodes "hostname; ethtool eth0 | grep Speed"

1.4.2 同步脚本，校对主机名，同步/etc/hosts

1 2 3 4 5 6 cluster_copy_all_nodes / root / .bash_profile / root / cluster_copy_all_nodes / usr / local / bin / cluster_run_all_nodes / usr / local / bin / cluster_copy_all_nodes / usr / local / bin / cluster_copy_all_nodes / usr / local / bin / cluster_run_all_nodes "hostname; /bin/hostname -f; grep HOSTNAME /etc/sysconfig/network" cluster_copy_all_nodes / etc / hosts / etc / cluster_run_all_nodes "hostname; cat /etc/hosts"

1.4.3 同步时间、时区、NTP服务状态

1 2 3 4 5 6 7 cluster_run_all_nodes 'hostname;date' cluster_run_all_nodes "date 032411082015.00" cluster_run_all_nodes "hwclock -r" cluster_run_all_nodes "hwclock -w" cluster_run_all_nodes "hostname; echo \${TZ}; echo \${LANG}" cluster_run_all_nodes "hostname; cat /etc/sysconfig/clock" cluster_run_all_nodes "hostname; /sbin/chkconfig --list ntpd"

1.4.4 同步Selinux配置，防火墙配置

1 2 3 4 5 6 7 cluster_run_all_nodes "hostname; grep 'SELINUX=' /etc/selinux/config" cluster_run_all_nodes "hostname; setenforce 0" vi / etc / selinux / config disabled cluster_copy_all_nodes / etc / selinux / config / etc / selinux / cluster_run_all_nodes "hostname; /sbin/chkconfig --list iptables" cluster_run_all_nodes "hostname; /sbin/chkconfig --level 0123456 iptables off" cluster_run_all_nodes "service iptables stop"

1.4.5 同步CPU、内存配置

1 2 3 4 5 cluster_run_all_nodes "hostname; grep processor /proc/cpuinfo | wc -l" cluster_run_all_nodes "hostname; grep MHz /proc/cpuinfo | sort -u" cluster_run_all_nodes "hostname; grep MemTotal /proc/meminfo" cluster_run_all_nodes "hostname; /sbin/chkconfig --list cpuspeed" cluster_run_all_nodes "/sbin/chkconfig --level 0123456 cpuspeed off;/sbin/service cpuspeed stop"

1.4.6 同步检查rsync、python版本

1 2	`cluster_run_all_nodes` `"hostname; rsync --version \| grep version"` `cluster_run_all_nodes` `"hostname; /usr/bin/python -V"`

1.4.7 同步IO配置

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 - - 选择<deadline> cluster_run_all_nodes "hostname; /sbin/modinfo cciss | grep version" cluster_run_all_nodes "hostname; cat /sys/block/sda/queue/scheduler" - - cluster_run_all_nodes "hostname; cat /sys/block/sdb/queue/scheduler" cluster_run_all_nodes "hostname; echo deadline > /sys/block/sda/queue/scheduler" - - cluster_run_all_nodes "hostname; echo deadline > /sys/block/sdb/queue/scheduler" cluster_run_all_nodes "echo 'deadline > /sys/block/sda/queue/scheduler' >> /etc/rc.d/rc.local" - - cluster_run_all_nodes "echo 'deadline > /sys/block/sdb/queue/scheduler' >> /etc/rc.d/rc.local" cluster_run_all_nodes "grep scheduler /etc/rc.d/rc.local" noop anticipatory [deadline] cfq - - 修改 / sys / kernel / mm / redhat_transparent_hugepage / enabled cluster_run_all_nodes "hostname; cat /sys/kernel/mm/redhat_transparent_hugepage/enabled" cluster_run_all_nodes "hostname; echo never > /sys/kernel/mm/redhat_transparent_hugepage/enabled" cluster_run_all_nodes "echo 'never > /sys/kernel/mm/redhat_transparent_hugepage/enabled' >> /etc/rc.d/rc.local" cluster_run_all_nodes "grep hugepage /etc/rc.d/rc.local" always [never] - - 修改 / sys / kernel / mm / redhat_transparent_hugepage / defrag cluster_run_all_nodes "hostname; cat /sys/kernel/mm/redhat_transparent_hugepage/defrag" cluster_run_all_nodes "hostname; echo never > /sys/kernel/mm/redhat_transparent_hugepage/defrag" cluster_run_all_nodes "echo 'never > /sys/kernel/mm/redhat_transparent_hugepage/defrag' >> /etc/rc.d/rc.local" cluster_run_all_nodes "grep hugepage /etc/rc.d/rc.local" always [never] - - 修改 / sys / kernel / mm / redhat_transparent_hugepage / khugepaged / defrag cluster_run_all_nodes "hostname; cat /sys/kernel/mm/redhat_transparent_hugepage/khugepaged/defrag" cluster_run_all_nodes "hostname; echo no > /sys/kernel/mm/redhat_transparent_hugepage/khugepaged/defrag" cluster_run_all_nodes "echo 'no > /sys/kernel/mm/redhat_transparent_hugepage/khugepaged/defrag' >> /etc/rc.d/rc.local" cluster_run_all_nodes "grep hugepage /etc/rc.d/rc.local" yes [no] - - 修改 / proc / sys / vm / swappiness cluster_run_all_nodes "hostname; cat /proc/sys/vm/swappiness" cluster_run_all_nodes "hostname; echo 0 > /proc/sys/vm/swappiness" cluster_run_all_nodes "echo '0 > /proc/sys/vm/swappiness' >> /etc/rc.d/rc.local" cluster_run_all_nodes "grep swappiness /etc/rc.d/rc.local" - - 修改readahead cluster_run_all_nodes "hostname; /sbin/blockdev --getra /dev/sda" cluster_run_all_nodes "hostname; /sbin/blockdev --setra 8192 /dev/sda" cluster_run_all_nodes "echo '/sbin/blockdev --setra 8192 /dev/sda' >> /etc/rc.d/rc.local" cluster_run_all_nodes "grep blockdev /etc/rc.d/rc.local"

1.4.8 同步系统配置

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 - - 同步 / etc / security / limits.conf vi / etc / security / limits.conf * soft nofile 655360 * hard nofile 655360 dbadmin - nice 0 dbadmin soft nproc 145209 dbadmin hard nproc 145209 cluster_run_all_nodes "hostname; ulimit -n 655360" cluster_run_all_nodes "hostname; ulimit -n" cluster_copy_all_nodes / etc / security / limits.conf / etc / security / - - 同步 / etc / sysctl.conf vm.max_map_count = 9293346 vm.min_free_kbytes = 65535 fs. file - max = 13226642 net.core.rmem_default = 262144 net.core.rmem_max = 262144 net.core.wmem_default = 262144 net.core.wmem_max = 262144 cluster_copy_all_nodes / etc / sysctl.conf / etc / cluster_run_all_nodes "hostname;sysctl -p /etc/sysctl.conf" - - 同步 / etc / pam.d / su session required pam_limits.so cluster_copy_all_nodes / etc / pam.d / su / etc / pam.d / cluster_run_all_nodes "hostname; grep session /etc/pam.d/su"

二、Vertica安装

1 2 3 rpm - ivh vertica - 7.1 . 0 - 3.x86_64 .RHEL5.rpm / opt / vertica / sbin / install_vertica - s DB01,DB02,DB03,DB04 - r / usr2 / vertica - 7.1 . 0 - 3.x86_64 .RHEL5.rpm - - failure - threshold = HALT - u dbadmin - p vertica

1 2	`cluster_run_all_nodes` `"hostname;mkdir -p /data/verticadb"` `cluster_run_all_nodes` `"hostname;chown -R dbadmin:verticadba /data/verticadb"`

三、集群性能评估

1 2 3 4 5 6 cluster_run_all_nodes "hostname; /opt/vertica/bin/vcpuperf" > /tmp/vcpuperf . log cluster_run_all_nodes "hostname; /opt/vertica/bin/vioperf /data" > /tmp/vioperf_data . log su - dbadmin /opt/vertica/bin/vnetperf > /tmp/vnetperf . log