Greenplum学习实践-【安装部署】-2、 5.10集群部署
-
Greenplum學習實踐-【安裝部署】-2、 5.10集群部署
?
?
第 1 章** 環境說明
1.1 官方文檔
Greenplum官方安裝說明:
https://gpdb.docs.pivotal.io/5160/install_guide/install_extensions.html
?
?
?
1.2 系統要求
?
| File Systems | ? xfs required for data storage on SUSE Linux and Red Hat (ext3 supported for root file system) |
| Minimum CPU | Pentium Pro compatible (P3/Athlon and above) |
?
| Disk Requirements | ? 150MB per host for Greenplum installation ? Approximately 300MB per segment instance for meta data ? Appropriate free space for data with disks at no more than 70% capacity ? High-speed, local storage |
| Network Requirements | 10 Gigabit Ethernet within the array Dedicated, non-blocking switch NIC bonding is recommended when multiple interfaces are present |
| Software and Utilities | zlib compression libraries bash shell GNU tars GNU zip GNU sed (used by Greenplum Database gpinitsystem) perl secure shell |
*Important**:* SSL is supported only on the Greenplum Database master host system.
?
1.3 搭建環境介紹
?
操作系統：CentOS Linux release 7.4.1708 (Core)
?
| CPU | Intel(R) Xeon(R) CPU E3-1220 v5 @ 3.00GHz *4物理核心（16cores） |
| 內存 | 16G |
| 磁盤 | 4T |
| Swap | 32G |
?
?
?
?
第 2 章 安裝架構
2.1 安裝大致步驟
Perform the following tasks in order: *1.* Make sure your systems meet the System Requirements *2.* Setting the Greenplum Recommended OS Parameters *3.* (master only) Creating the Greenplum Database Administrative User Account *4.* (master only) Installing the Greenplum Database Software *5.* Installing and Configuring Greenplum on all Hosts *6.* (Optional) Installing Oracle Compatibility Functions *7.* (Optional) Installing Optional Modules *8.* (Optional) Installing Greenplum Database Extensions *9.* (Optional) Installing and Configuring the Greenplum Platform Extension Framework (PXF) *10.*Creating the Data Storage Areas *11.*Synchronizing System Clocks *12.*Next Steps
?
2.2 軟件信息
greenplum-db-5.10.2-rhel6-x86_64.zip是mpp軟件，
greenplum-cc-web-4.3.1-LINUX-x86_64.zip是web監控平臺
?
2.3 架構說明
準備 4 臺服務器， 1 臺做 master， 3臺都做存儲共部署 6 個segment 及其鏡像
| 10.102.254.24 | sdw1 | 16 | 16 | 4*segment |
| 10.102.254.25 | sdw2 | 16 | 16 | 4*segment |
| 10.102.254.26 | sdw3 | 16 | 16 | 4*segment |
| 10.102.254.27 | mdw1 | 16 | 16 | master |
?
=架構目標==
?
| master | seg0p seg1p seg5m seg4m | seg2p seg3p seg0m seg1m | seg4p seg5p seg2m seg3m smdw |
?
第 3 章 安裝架構
3.1 關閉防火墻和selinux-所有節點
?systemctl stop firewalld.service ?systemctl disable firewalld.service ?iptables -F ?vi /etc/selinux/config SELINUX=DISABLED?# sestatus SELinuxstatus: disabled ? ???
3.2 修改host文件 -所有節點
?
?
?
?
?
3.2 修改主機名
?
?hostnamectl set-hostname sdw1?hostnamectl status 狀態?/etc/sysconfig/network???
3.3 系統內核參數優化
?
???vi /etc/sysctl.conf?kernel.shmmax = 500000000?kernel.shmmni = 4096?kernel.shmall = 4000000000?kernel.sem = 250 512000 100 2048?kernel.sysrq = 1?kernel.core_uses_pid = 1?kernel.msgmnb = 65536?kernel.msgmax = 65536?kernel.msgmni = 2048?net.ipv4.tcp_syncookies = 1?net.ipv4.ip_forward = 0?net.ipv4.conf.default.accept_source_route = 0?net.ipv4.tcp_tw_recycle = 1?net.ipv4.tcp_max_syn_backlog = 4096?net.ipv4.conf.all.arp_filter = 1?net.ipv4.ip_local_port_range = 1025 65535?net.core.netdev_max_backlog = 10000?net.core.rmem_max = 2097152?net.core.wmem_max = 2097152?vm.overcommit_memory = 2 ??sysctl -p???cat > /etc/sysctl.conf << EOF?\# sysctl settings are defined through files in?\# /usr/lib/sysctl.d/, /run/sysctl.d/, and /etc/sysctl.d/.?\#?\# Vendors settings live in /usr/lib/sysctl.d/.?\# To override a whole file, create a new file with the same in?\# /etc/sysctl.d/ and put new settings there. To override?\# only specific settings, add a file with a lexically later?\# name in /etc/sysctl.d/ and put new settings there.?\#?\# For more information, see sysctl.conf(5) and sysctl.d(5).??kernel.shmmax = 500000000?kernel.shmmni = 4096?kernel.shmall = 4000000000?kernel.sem = 500 1024000 200 4096?kernel.sysrq = 1?kernel.core_uses_pid = 1?kernel.msgmnb = 65536?kernel.msgmax = 65536?kernel.msgmni = 2048?net.ipv4.tcp_syncookies = 1?net.ipv4.ip_forward = 0?net.ipv4.conf.default.accept_source_route = 0?net.ipv4.tcp_tw_recycle = 1?net.ipv4.tcp_max_syn_backlog = 4096?net.ipv4.conf.all.arp_filter = 1?net.ipv4.ip_local_port_range = 1025 65535?net.core.netdev_max_backlog = 10000?net.core.rmem_max = 2097152?net.core.wmem_max = 2097152?vm.overcommit_memory = 2?vm.swappiness = 1?kernel.pid_max = 655350?EOF??sysctl -p?
3.4 修改Linux最大限制
?
?cat /etc/security/limits.conf??vi /etc/security/limits.conf??* soft nofile 65536?* hard nofile 65536?* soft nproc 131072?* hard nproc 131072??cat > /etc/security/limits.conf << EOF?* soft nofile 65536?* hard nofile 65536?* soft nproc 131072?* hard nproc 131072?EOF?
如何是rhel 6.x 請注意 /etc/security/limits.d/90-nproc.conf，詳細情況請見文檔
?
3.5 設備與IO-文件系統
設置XFS文件系統并掛載
EXT4是第四代擴展文件系統（英語：Fourth EXtended filesystem，縮寫為ext4）是Linux系統下的日志文件系統，是ext3文件系統的后繼版本。
Ext4的文件系統容量達到1EB，而文件容量則達到16TB，這是一個非常大的數字了。對一般的臺式機和服務器而言，這可能并不重要，但對于大型磁盤陣列的用戶而言，這就非常重要了。
XFS是一個64位文件系統，最大支持8EB減1字節的單個文件系統，實際部署時取決于宿主操作系統的最大塊限制。對于一個32位Linux系統，文件和文件系統的大小會被限制在16TB。
?
二者各有特點，而性能表現基本上是差不多的。例如，谷歌公司就考慮將EXT2系統升級，最終確定為EXT4系統。谷歌公司表示，他們還考慮過XFS和JFS。結果顯示，EXT4和XFS的表現類似，不過從EXT2升級到EXT4比升級到XFS容易。
?
例子：
cat >> /etc/fstab << EOF
/dev/sdb1 /greenplum xfs rw,nodev,noatime,inode64,allocsize=16m 0 0
EOF
rw,nodev,noatime,nobarrier,inode64
?
cat /etc/fstab
?
?
?
3.6 磁盤訪問策略
Linux磁盤I/O調度器對磁盤的訪問支持不同的策略，默認的為CFQ，GP建議設置為deadline
?
查看磁盤的I/O調度策略，看到默認的為[cfq]
?
The deadline scheduler option is recommended. To specify a scheduler until the next system reboot,
run the following:
# echo schedulername > /sys/block/devname/queue/scheduler
?
echo deadline > /sys/block/sda/queue/scheduler
?
linux 7
# grubby --update-kernel=ALL --args="elevator=deadline"
grubby --info=ALL
?
?
3.7 調整磁盤預讀扇區數
?
fdisk -l
?
檢查
/sbin/blockdev --getra /dev/sda
設置
/sbin/blockdev --setra 16384 /dev/sda
?
?
在參數文件 /etc/rc.d/rc.local 中增加
DELL : blockdev --setra 16384 /dev/sd* (紅色部分為硬盤設備標識) HP:blockdev --setra 16384 /dev/cciss/c?d?*
?
?
3.8 禁用THP
On systems that use grub2 such as RHEL 7.x or CentOS 7.x, use the system utility grubby. This
command adds the parameter when run as root.
# grubby --update-kernel=ALL --args="transparent_hugepage=never"
After adding the parameter, reboot the system.
This cat command checks the state of THP. The output indicates that THP is disabled.
$ cat /sys/kernel/mm/*transparent_hugepage/enabled
always [never]
?
服務方式注冊
# 創建 init.d 腳本
echo '#!/bin/sh
case $1 in
start)
if [ -d /sys/kernel/mm/transparent_hugepage ]; then
thp_path=/sys/kernel/mm/transparent_hugepage
elif [ -d /sys/kernel/mm/redhat_transparent_hugepage ]; then
thp_path=/sys/kernel/mm/redhat_transparent_hugepage
else
exit 0
fi
?
echo never > ${thp_path}/enabled
echo never > ${thp_path}/defrag
?
unset thp_path
;;
esac' > /etc/init.d/disable-transparent-hugepages
?
# 注冊systemd文件
echo '[Unit]
Description=Disable Transparent Hugepages
After=multi-user.target
?
[Service]
ExecStart=/etc/init.d/disable-transparent-hugepages start
Type=simple
?
[Install]
WantedBy=multi-user.target' > /etc/systemd/system/disable-thp.service
?
# 磁盤預讀扇區數
/sbin/blockdev --getra /dev/sdb1 # 查看大小
/sbin/blockdev --setra 65535 /dev/sdb1 # 設置大小
?
# 創建 init.d 腳本
echo '#!/bin/sh
device_name=/dev/sdb1
case $1 in
start)
if mount | grep "^${device_name}" > /dev/null;then
/sbin/blockdev --setra 65535 ${device_name}
else
exit 0
fi
?
unset device_name
;;
esac' > /etc/init.d/blockdev-setra-sdb
?
# 注冊systemd文件
echo '[Unit]
Description=Blocdev --setra N
After=multi-user.target
?
[Service]
ExecStart=/etc/init.d/blockdev-setra-sdb start
Type=simple
?
[Install]
WantedBy=multi-user.target' > /etc/systemd/system/blockdev-setra-sdb.service
?
# 授權并設置開機啟動
chmod 755 /etc/init.d/disable-transparent-hugepages
chmod 755 /etc/init.d/blockdev-setra-sdb
chmod 755 /etc/systemd/system/disable-thp.service
chmod 755 /etc/systemd/system/blockdev-setra-sdb.service
systemctl enable disable-thp blockdev-setra-sdb
?
?
?
3.9 Disable IPC object removal for RHEL 7 or CentOS 7
?
Set this parameter in /etc/systemd/logind.conf on the Greenplum
Database host systems.
RemoveIPC=no
The setting takes effect after restarting the systemd-login service or rebooting the system. To
restart the service, run this command as the root user.
service systemd-logind restart
?
cat /etc/systemd/logind.conf
?
3.10 時間同步
?
/etc/chrony.conf
?
systemctl status chronyd.service --查看狀態
systemctl start chronyd.service --啟動
systemctl enable chronyd.service --使其開機自啟
systemctl status chronyd.service
server 10.1.3.1 prefer
?
查看時間同步源
chronyc sources -v
chronyc sourcestats -v
?
3.11 控制ssh連接數
/etc/ssh/sshd_config
MaxStartups 10:30:200
?
systemctl restart sshd.service
?
?
?
3.12 系統依賴包
yum -y install epel-release
yum -y install wget cmake3 git gcc gcc-c++ bison flex libedit-devel zlib zlib-devel perl-devel perl-ExtUtils-Embed python-devel libevent libevent-devel libxml2 libxml2-devel libcurl libcurl-devel bzip2 bzip2-devel net-tools libffi-devel openssl-devel
?
第 4章 安裝軟件
4.1 創建用戶和組
master only
?# groupadd gpadmin?# useradd gpadmin -g gpadmin?# passwd gpadmin?New password: <changeme>?Retype new password: <changeme>???echo gpadmin | passwd gpadmin --stdin?
?
?
4.2 root用戶解壓縮和安裝
?
?./greenplum-db-5.10.2-rhel6-x86_64.bin???I HAVE READ AND AGREE TO THE TERMS OF THE ABOVE PIVOTAL SOFTWARE?LICENSE AGREEMENT.?????********************************************************************************?Do you accept the Pivotal Database license agreement? [yes|no]?********************************************************************************???yes???********************************************************************************?Provide the installation path for Greenplum Database or press ENTER to?accept the default installation path: /usr/local/greenplum-db-5.10.2?********************************************************************************???????********************************************************************************?Install Greenplum Database into /usr/local/greenplum-db-5.10.2? [yes|no]?********************************************************************************???yes???********************************************************************************?/usr/local/greenplum-db-5.10.2 does not exist.?Create /usr/local/greenplum-db-5.10.2 ? [yes|no]?(Selecting no will exit the installer)?********************************************************************************???????安裝完成后授權?????\# chown -R gpadmin /usr/local/greenplum*（在創建gpadmin后執行）???\# chgrp -R gpadmin /usr/local/greenplum*（在創建gpadmin后執行）?
?
4.3 編輯環境變量
?
?cat >> .bashrc << EOF?export MASTER_DATA_DIRECTORY=/greenplum/gpdata/master/gpseg-1?source /usr/local/greenplum-db/greenplum_path.sh?EOF?source .bashrc?????cat >> /home/gpadmin/.bash_profile <<EOF ??export MASTER_DATA_DIRECTORY=/greenplum/gpdata/master/gpseg-1?source /usr/local/greenplum-db/greenplum_path.sh?export PGPORT=5432?export PGDATABASE=archdata?EOF???source /home/gpadmin/.bash_profile?
?
?
4.4 進行文件配置
?
切換root
?
source /usr/local/greenplum-db/greenplum_path.sh
?
?
?
------只在mdw，smdw執行
mkdir /home/gpadmin/gpconfig
chown -R gpadmin:gpadmin /home/gpadmin/gpconfig
?
------只在mdw，smdw執行
cat >> /home/gpadmin/gpconfig/all_host <<EOF
mdw
sdw1
sdw2
sdw3
EOF
?
------只在mdw，smdw執行
cat >> /home/gpadmin/gpconfig/all_segment <<EOF
sdw1
sdw2
sdw3
EOF
?
chown -R gpadmin:gpadmin /home/gpadmin/gpconfig/all_host
chown -R gpadmin:gpadmin /home/gpadmin/gpconfig/all_segment
?
4.5 設置主機免密碼登陸 -
?
source /usr/local/greenplum-db/greenplum_path.sh
?
/usr/local/greenplum-db/bin/gpssh-exkeys -f /home/gpadmin/gpconfig/all_host
?
4.6 確認檢查主機連接狀態
?
gpssh -f /home/gpadmin/gpconfig/all_host -e "ls -l"
?
?
4.7 批量創建其他節點的用戶
?
gpssh -f /home/gpadmin/gpconfig/all_segment
?
groupadd gpadmin
useradd gpadmin -g gpadmin
passwd gpadmin
echo gpadmin | passwd gpadmin --stdin
?
?
4.8 gpadmin用戶-互信
?
source /usr/local/greenplum-db/greenplum_path.sh
?
/usr/local/greenplum-db/bin/gpssh-exkeys -f /home/gpadmin/gpconfig/all_host
?
gpssh -f /home/gpadmin/gpconfig/all_host -e "ls -l"
?
4.9 檢查時間同步
gpssh -f /home/gpadmin/gpconfig/all_host -e "date"
?
4.10 分發所有seg節點軟件
?
root執行
?
source /usr/local/greenplum-db/greenplum_path.sh
?
gpseginstall -f /home/gpadmin/gpconfig/all_host -u gpadmin -p gpadmin
?
?
4.11 檢查安裝情況
?
o Log in as the gpadmin user and source
? source /usr/local/greenplum-db/greenplum_path.sh
o Use the gpssh utility to see if you can login to all hosts without a password prompt
?
?
4.12 創建相關目錄(root用戶)
?
mkdir -p /greenplum/gpdata/master
chown gpadmin:gpadmin /greenplum/gpdata/master
?
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'mkdir -p /greenplum/gpdata/primary1'
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'mkdir -p /greenplum/gpdata/primary2'
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'chown -R gpadmin:gpadmin /greenplum/gpdata'
?
?
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'mkdir -p /greenplum/gpdata/mirror1'
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'mkdir -p /greenplum/gpdata/mirror2'
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'chown -R gpadmin:gpadmin /greenplum/gpdata'
?
or 批量創建
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'mkdir -p /greenplum/gpdata/primary{1..2}'
gpssh -f /home/gpadmin/gpconfig/all_segment -e 'chown -R gpadmin:gpadmin /greenplum/gpdata'
?
4.13 驗證系統
?
檢查系統參數和測試性能
檢查命令：gpcheck -f host_file -m mdw -ssmdw
Validating Hardware Performance
o gpcheckperf can be used to identify hardware and system-level issues on the machines in your Greenplum
Database array.
o Network Performance (gpnetbench*)
? gpcheckperf -f hostfile_gpchecknet_ic1 -r N -d /tmp > subnet1.out
o Disk I/O Performance (dd test) & Memory Bandwidth (stream test)
? gpcheckperf -f hostfile_gpcheckperf -r ds -D -d /data/primary -d /data/mirror
?
?
4.14 驗證OS配置
?
source /usr/local/greenplum-db/greenplum_path.sh
gpcheck -f /home/gpadmin/gpconfig/all_host -m mdw
?
?
驗證硬件性能--這個需要確認（網絡和IO）
?
? gpcheckperf -f /home/gpadmin/gpconfig/all_host -r N -d /tmp > checknetwork.out?[root@mdw greenplum-db]# cat checknetwork.out?/usr/local/greenplum-db/./bin/gpcheckperf -f /home/gpadmin/gpconfig/all_host -r N -d /tmp???\-------------------?-- NETPERF TEST?\-------------------???====================?== RESULT?====================?Netperf bisection bandwidth test?mdw -> sdw1 = 112.340000?sdw2 -> sdw3 = 112.340000?sdw1 -> mdw = 112.330000?sdw3 -> sdw2 = 112.330000???Summary:?sum = 449.34 MB/sec?min = 112.33 MB/sec?max = 112.34 MB/sec?avg = 112.33 MB/sec?median = 112.34 MB/sec???gpcheckperf -f /home/gpadmin/gpconfig/all_host -r ds -D -d /greenplum/gpdata/primary1 -d /greenplum/gpdata/mirror1 > checkDISKIO.out?[root@mdw greenplum-db]# gpcheckperf -f /home/gpadmin/gpconfig/all_host -r ds -D -d /greenplum/gpdata/primary1 -d /greenplum/gpdata/mirror1?/usr/local/greenplum-db/./bin/gpcheckperf -f /home/gpadmin/gpconfig/all_host -r ds -D -d /greenplum/gpdata/primary1 -d /greenplum/gpdata/mirror1???\--------------------?-- DISK WRITE TEST?\--------------------???\--------------------?-- DISK READ TEST?\--------------------???\--------------------?-- STREAM TEST?\--------------------???====================?== RESULT?====================???disk write avg time (sec): 20.88?disk write tot bytes: 132920115200?disk write tot bandwidth (MB/s): 6074.65?disk write min bandwidth (MB/s): 1476.04 [ mdw]?disk write max bandwidth (MB/s): 1551.18 [sdw3]?-- per host bandwidth --?disk write bandwidth (MB/s): 1476.04 [ mdw]?disk write bandwidth (MB/s): 1537.63 [sdw1]?disk write bandwidth (MB/s): 1509.80 [sdw2]?disk write bandwidth (MB/s): 1551.18 [sdw3]?????disk read avg time (sec): 59.80?disk read tot bytes: 132920115200?disk read tot bandwidth (MB/s): 2175.57?disk read min bandwidth (MB/s): 454.54 [sdw2]?disk read max bandwidth (MB/s): 700.04 [sdw1]?-- per host bandwidth --?disk read bandwidth (MB/s): 520.03 [ mdw]?disk read bandwidth (MB/s): 700.04 [sdw1]?disk read bandwidth (MB/s): 454.54 [sdw2]?disk read bandwidth (MB/s): 500.96 [sdw3]?????stream tot bandwidth (MB/s): 49348.52?stream min bandwidth (MB/s): 12297.76 [ mdw]?stream max bandwidth (MB/s): 12388.57 [sdw2]?-- per host bandwidth --?stream bandwidth (MB/s): 12297.76 [ mdw]?stream bandwidth (MB/s): 12321.47 [sdw1]?stream bandwidth (MB/s): 12388.57 [sdw2]?stream bandwidth (MB/s): 12340.73 [sdw3]?
第 5 章 初始化database
copy配置文件
?
設置初始化系統文件列表
?
調整參數
?
?
初始化database
?
?
gpadmin用戶
?
?gpinitsystem -c /home/gpadmin/gpconfig/gpinitsystem_config -h /home/gpadmin/gpconfig/hostfile_gpinitsystem???如何添加master standby和修改mirror分布策略spread mirror?gpinitsystem -c gpconfigs/gpinitsystem_config -h gpconfigs/hostfile_gpinitsystem -s?standby_master_hostname -S (with a standby master and a spread mirror configuration)?
檢查環境變量
?
???MASTER_DATA_DIRECTORY=/data/master/gpseg-1???GPHOME=/usr/local/greenplum-db???PGDATABASE=gpadmin???[gpadmin@mdw ~]$ cat .bash_profile?\# .bash_profile???\# Get the aliases and functions?if [ -f ~/.bashrc ]; then?? ? . ~/.bashrc?fi???\# User specific environment and startup programs???PATH=$PATH:$HOME/.local/bin:$HOME/bin???export PATH?export MASTER_DATA_DIRECTORY=/greenplum/gpdata/master/gpseg-1?source /usr/local/greenplum-db/greenplum_path.sh?export PGPORT=5432?export PGDATABASE=archdata???
第 6 章 連接測試
?
?
6.1設置gpadmin遠程訪問密碼
psql postgres gpadmin
alter user gpadmin encrypted password 'gpadmin';
\q
?
6.2查詢測試
psql -hmdw -p 5432 -d postgres -U gpadmin -c 'select dfhostname, dfspace,dfdevice from gp_toolkit.gp_disk_free order by dfhostname;'
?
[gpadmin@mdw ~]$ psql -hmdw -p 5432 -d postgres -U gpadmin -c 'select dfhostname, dfspace,dfdevice from gp_toolkit.gp_disk_free order by dfhostname;'
dfhostname | dfspace | dfdevice
------------+----------+----------------------------
sdw1 | 98708120 | /dev/mapper/VolGroup-root
sdw1 | 98708120 | /dev/mapper/VolGroup-root
sdw2 | 98705600 | /dev/mapper/VolGroup-root
sdw2 | 98705600 | /dev/mapper/VolGroup-root
sdw3 | 98705144 | /dev/mapper/VolGroup-root
sdw3 | 98705144 | /dev/mapper/VolGroup-root
(6 rows)
?
psql -h hmdw -p 5432 -d postgres -U gpadmin -c '\l+'
[gpadmin@mdw ~]$ psql -h mdw -p 5432 -d postgres -U gpadmin -c '\l+'
List of databases
Name | Owner | Encoding | Access privileges | Size | Tablespace | Description
-----------+---------+----------+---------------------+-------+------------+---------------------------
postgres | gpadmin | UTF8 | | 73 MB | pg_default |
template0 | gpadmin | UTF8 | =c/gpadmin | 72 MB | pg_default |
: gpadmin=CTc/gpadmin
template1 | gpadmin | UTF8 | =c/gpadmin | 73 MB | pg_default | default template database
: gpadmin=CTc/gpadmin
(3 rows)
?
[gpadmin@mdw ~]$
?
第 7 章 常用命令
?
數據庫啟動：gpstart 常用可參數： -a : 直接啟動，不提示終端用戶輸入確認 -m:只啟動master 實例，主要在故障處理時使用
數據庫停止：gpstop： 常用可參數：-a：直接停止，不提示終端用戶輸入確認 -m：只停止master 實例，與gpstart –m 對應使用 -M fast：停止數據庫，中斷所有數據庫連接，回滾正在運 行的事務 -u：不停止數據庫，只加載pg_hba.conf 和postgresql.conf中運行時參數，當改動參數配置時候使用。 評：-a用在shell里，最多用的還是-M fast。
查看實例配置和狀態 select * from gp_configuration order by 1 ; 主要字段說明： Content：該字段相等的兩個實例，是一對Ｐ（primary instance）和Ｍ（mirror Instance) Isprimary：實例是否作為primary instance 運行 Valid：實例是否有效，如處于false 狀態，則說明該實例已經down 掉。 Port：實例運行的端口 Datadir:實例對應的數據目錄
gpstate ：顯示Greenplum數據庫運行狀態，詳細配置等信息 常用可參數：-c：primary instance 和 mirror instance 的對應關系 -m：只列出mirror 實例的狀態和配置信息 -f：顯示standby master 的詳細信息 -Q：顯示狀態綜合信息 該命令默認列出數據庫運行狀態匯總信息，常用于日常巡檢。 評：最開始由于網卡驅動的問題，做了mirror后，segment經常down掉，用-Q參數查詢綜合信息還是比較有用的。
查看用戶會話和提交的查詢等信息 select * from pg_stat_activity 該表能查看到當前數據庫連接的IP 地址，用戶名，提交的查詢等。另外也可以在master 主機上查看進程，對每個客戶端連接，master 都會創建一個進程。ps -ef |grep -i postgres |grep -i con 評：常用的命令，我經常用這個查看數據庫死在那個sql上了。
查看數據庫、表占用空間 select pg_size_pretty(pg_relation_size('schema.tablename')); select pg_size_pretty(pg_database_size('databasename')); 必須在數據庫所對應的存儲系統里，至少保留30%的自由空間，日常巡檢，要檢查存儲空間的剩余容量。 評：可以查看任何數據庫對象的占用空間，pg_size_pretty可以顯示如mb之類的易讀數據，另外，可以與pg_tables，pg_indexes之類的系統表鏈接，統計出各類關于數據庫對象的空間信息。
收集統計信息，回收空間 定期使用Vacuum analyze tablename 回收垃圾和收集統計信息，尤其在大數據量刪除，導入以后，非常重要 評：這個說的不全面，vacuum分兩種，一種是analize，優化查詢計劃的，還有一種是清理垃圾數據，postres刪除工作，并不是真正刪除數據，而是在被刪除的數據上，坐一個標記，只有執行vacuum時，才會真正的物理刪除，這個非常重用，有些經常更新的表，各種查詢、更新效率會越來越慢，這個多是因為沒有做vacuum的原因。
查看數據分布情況 兩種方式： l Select gp_segment_id,count(*) from tablename group by 1 ; l 在命令運行：gpskew -t public.ate -a postgres 如數據分布不均勻，將發揮不了并行計算的優勢，嚴重影響性能。 評：非常用，gp要保障數據分布均勻。
實例恢復：gprecoverseg 通過gpstate 或gp_configuration 發現有實例down 掉以后，使用該命令進行回復。
查看鎖信息： SELECT locktype, database, c.relname, l.relation, l.transactionid, l.transaction, l.pid, l.mode, l.granted, a.current_query FROM pg_locks l, pg_class c, pg_stat_activity a WHERE l.relation=c.oid AND l.pid=a.procpid ORDER BY c.relname; 主要字段說明： relname: 表名 locktype、mode 標識了鎖的類型
explain：在提交大的查詢之前，使用explain分析執行計劃、發現潛在優化機會，避免將系統資源熬盡。 評：少寫了個analyze，如果只是explain，統計出來的執行時間，是非常坑爹的，如果希望獲得準確的執行時間，必須加上analyze。
數據庫備份 gp_dump 常用參數：-s: 只導出對象定義（表結構，函數等） -n: 只導出某個schema gp_dump 默認在master 的data 目錄上產生這些文件： gp_catalog_1__ ：關于數據庫系統配置的備份文件 gp_cdatabase_1：數據庫創建語句的備份文件 gp_dump_1：數據庫對象ddl語句 gp_dump_status_1：備份操作的日志 在每個segment instance 上的data目錄上產生的文件： gp_dump_0：用戶數據備份文件 gp_dump_status_0：備份日志
數據庫恢復 gp_restore 必參數：--gp-k=key ：key 為gp_dump 導出來的文件的后綴時間戳 -d dbname ：將備份文件恢復到dbname
登陸與退出Greenplum #正常登陸 psql gpdb psql -d gpdb -h gphostm -p 5432 -U gpadmin #使用utility方式 PGOPTIONS="-c gp_session_role=utility" psql -h -d dbname hostname -p port #退出 在psql命令行執行\q
參數查詢 psql -c 'SHOW ALL;' -d gpdb gpconfig --show max_connections 評：這個用，可以管道給grep。
創建數據庫 createdb -h localhost -p 5432 dhdw
創建GP文件系統 # 文件系統名 gpfsdw # 子節點，視segment數創建目錄 mkdir -p /gpfsdw/seg1 mkdir -p /gpfsdw/seg2 chown -R gpadmin:gpadmin /gpfsdw # 主節點 mkdir -p /gpfsdw/master chown -R gpadmin:gpadmin /gpfsdw gpfilespace -o gpfilespace_config gpfilespace -c gpfilespace_config
創建GP表空間 psql gpdb create tablespace TBS_DW_DATA filespace gpfsdw; SET default_tablespace = TBS_DW_DATA;
刪除GP數據庫 gpdeletesystem -d /gpmaster/gpseg-1 -f
查看segment配置 select * from gp_segment_configuration;
文件系統 select * from pg_filespace_entry;
磁盤、數據庫空間 SELECT * FROM gp_toolkit.gp_disk_free ORDER BY dfsegment; SELECT * FROM gp_toolkit.gp_size_of_database ORDER BY sodddatname; 日志 SELECT * FROM gp_toolkit.gp_log_master_ext; SELECT * FROM gp_toolkit.gp_log_segment_ext;
表數據分布 SELECT gp_segment_id, count(*) FROM GROUP BY gp_segment_id;
表占用空間 SELECT relname as name, sotdsize/1024/1024 as size_MB, sotdtoastsize as toast, sotdadditionalsize as other FROM gp_toolkit.gp_size_of_table_disk as sotd, pg_class WHERE sotd.sotdoid = pg_class.oid ORDER BY relname;
索引占用空間 SELECT soisize/1024/1024 as size_MB, relname as indexname FROM pg_class, gp_toolkit.gp_size_of_index WHERE pg_class.oid = gp_size_of_index.soioid AND pg_class.relkind='i';
OBJECT的操作統計 SELECT schemaname as schema, objname as table, usename as role, actionname as action, subtype as type, statime as time FROM pg_stat_operations WHERE objname = '';
鎖 SELECT locktype, database, c.relname, l.relation, l.transactionid, l.transaction, l.pid, l.mode, l.granted, a.current_query FROM pg_locks l, pg_class c, pg_stat_activity a WHERE l.relation=c.oid AND l.pid=a.procpid ORDER BY c.relname;
隊列 SELECT * FROM pg_resqueue_status;
gpfdist外部表 # 啟動服務 gpfdist -d /share/txt -p 8081 –l /share/txt/gpfdist.log & # 創建外部表，分隔符為’/t’ drop EXTERNAL TABLE TD_APP_LOG_BUYER; CREATE EXTERNAL TABLE TD_APP_LOG_BUYER ( IP text, ACCESSTIME text, REQMETHOD text, URL text, STATUSCODE int, REF text, name text, VID text) LOCATION ('gpfdist://gphostm:8081/xxx.txt') FORMAT 'TEXT' (DELIMITER E'/t' FILL MISSING FIELDS) SEGMENT REJECT LIMIT 1 percent;
創建普通表 create table test select * from TD_APP_LOG_BUYER; # 索引 # CREATE INDEX idx_test ON test USING bitmap (ip); # 查詢數據 select ip , count() from test group by ip order by count(); gpload # 創建控制文件 # 加載數據 gpload -f my_load.yml copy COPY country FROM '/data/gpdb/country_data' WITH DELIMITER '|' LOG ERRORS INTO err_country SEGMENT REJECT LIMIT 10 ROWS;
gpfdist外部表 創建可寫外部表 CREATE WRITABLE EXTERNAL TABLE unload_expenses ( LIKE expenses ) LOCATION ('gpfdist://etlhost-1:8081/expenses1.out', 'gpfdist://etlhost-2:8081/expenses2.out') FORMAT 'TEXT' (DELIMITER ',') DISTRIBUTED BY (exp_id); # 寫權限 GRANT INSERT ON writable_ext_table TO ; # 寫數據 INSERT INTO writable_ext_table SELECT * FROM regular_table; copy COPY (SELECT * FROM country WHERE country_name LIKE 'A%') TO '/home/gpadmin/a_list_countries.out'; 執行sql文件 psql gpdbname –f yoursqlfile.sql 或者psql登陸后執行 \i yoursqlfile.sql
總結
以上是生活随笔為你收集整理的Greenplum学习实践-【安装部署】-2、 5.10集群部署的全部內容，希望文章能夠幫你解決所遇到的問題。
- 上一篇： Win10修改远程桌面3389端口
- 下一篇： 因为热爱，所以坚持；因为坚持，得以突破！