注册
登录

您现在的位置是:首页 > 学无止境

PostgreSQL on Linux 最佳部署手册

木木彡82 2017-04-30 21:45:32 353人围观
转载自:http://mp.weixin.qq.com/s/FR65pyRmpEFFVvoJ28uBUg

背景

数据库的安装一直以来都挺复杂的,特别是Oracle,现在身边都还有安装Oracle数据库赚外快的事情。
PostgreSQL其实安装很简单,但是那仅仅是可用,并不是好用。很...

转载自:http://mp.weixin.qq.com/s/FR65pyRmpEFFVvoJ28uBUg

背景

数据库的安装一直以来都挺复杂的,特别是Oracle,现在身边都还有安装Oracle数据库赚外快的事情。

PostgreSQL其实安装很简单,但是那仅仅是可用,并不是好用。很多用户使用默认的方法安装好数据库之后,然后测试一通性能,发现性能不行就不用了。

原因不用说,多方面没有优化的结果。

PostgreSQL数据库为了适应更多的场景能使用,默认的参数都设得非常保守,通常需要优化,比如检查点,SHARED BUFFER等。

本文将介绍一下PostgreSQL on Linux的最佳部署方法,其实在我的很多文章中都有相关的内容,但是没有总结成一篇文档。

安装常用包

# yum -y install coreutils glib2 lrzsz mpstat dstat sysstat e4fsprogs xfsprogs ntp readline-devel zlib-devel openssl-devel pam-devel libxml2-devel libxslt-devel python-devel tcl-devel gcc make smartmontools flex bison perl-devel perl-ExtUtils* openldap-devel

配置OS内核参数

# vi /etc/sysctl.conf
 
# add by digoal.zhou
fs.aio-max-nr = 1048576
fs.file-max = 76724600
kernel.core_pattern= /data01/corefiles/core_%e_%u_%t_%s.%p         
# /data01/corefiles事先建好,权限777
kernel.sem = 4096 2147483647 2147483646 512000    
# 信号量, ipcs -l  -u 查看,每16个进程一组,每组信号量需要17个信号量。
kernel.shmall = 107374182      
# 所有共享内存段相加大小限制(建议内存的80%)
kernel.shmmax = 274877906944   
# 最大单个共享内存段大小(建议为内存一半), >9.2的版本已大幅降低共享内存的使用
kernel.shmmni = 819200         
# 一共能生成多少共享内存段,每个PG数据库集群至少2个共享内存段
net.core.netdev_max_backlog = 10000
net.core.rmem_default = 262144       
# The default setting of the socket receive buffer in bytes.
net.core.rmem_max = 4194304          
# The maximum receive socket buffer size in bytes
net.core.wmem_default = 262144       
# The default setting (in bytes) of the socket send buffer.
net.core.wmem_max = 4194304          
# The maximum send socket buffer size in bytes.
net.core.somaxconn = 4096
net.ipv4.tcp_max_syn_backlog = 4096
net.ipv4.tcp_keepalive_intvl = 20
net.ipv4.tcp_keepalive_probes = 3
net.ipv4.tcp_keepalive_time = 60
net.ipv4.tcp_mem = 8388608 12582912 16777216
net.ipv4.tcp_fin_timeout = 5
net.ipv4.tcp_synack_retries = 2
net.ipv4.tcp_syncookies = 1    
# 开启SYN Cookies。当出现SYN等待队列溢出时,启用cookie来处理,可防范少量的SYN攻击
net.ipv4.tcp_timestamps = 1    
# 减少time_wait
net.ipv4.tcp_tw_recycle = 0    
# 如果=1则开启TCP连接中TIME-WAIT套接字的快速回收,但是NAT环境可能导致连接失败,建议服务端关闭它
net.ipv4.tcp_tw_reuse = 1      
# 开启重用。允许将TIME-WAIT套接字重新用于新的TCP连接
net.ipv4.tcp_max_tw_buckets = 262144
net.ipv4.tcp_rmem = 8192 87380 16777216
net.ipv4.tcp_wmem = 8192 65536 16777216
net.nf_conntrack_max = 1200000
net.netfilter.nf_conntrack_max = 1200000
vm.dirty_background_bytes = 409600000       
系统脏页到达这个值,系统后台刷脏页调度进程 pdflush(或其他)自动将(dirty_expire_centisecs/100)秒前的脏页刷到磁盘
vm.dirty_expire_centisecs = 3000             
比这个值老的脏页,将被刷到磁盘。3000表示30秒。
vm.dirty_ratio = 95                          
如果系统进程刷脏页太慢,使得系统脏页超过内存 95 % 时,则用户进程如果有写磁盘的操作(如fsync, fdatasync等调用),则需要主动把系统脏页刷出。
有效防止用户进程刷脏页,在单机多实例,并且使用CGROUP限制单实例IOPS的情况下非常有效。  
vm.dirty_writeback_centisecs = 100            
#  pdflush(或其他)后台刷脏页进程的唤醒间隔, 100表示1秒。
vm.extra_free_kbytes = 4096000
vm.min_free_kbytes = 2097152
vm.mmap_min_addr = 65536
vm.overcommit_memory = 0     
在分配内存时,允许少量over malloc, 如果设置为 1, 则认为总是有足够的内存,内存较少的测试环境可以使用 1 .  
vm.overcommit_ratio = 90     
overcommit_memory = 2 时,用于参与计算允许指派的内存大小。
vm.swappiness = 0            
关闭交换分区
vm.zone_reclaim_mode = 0     
# 禁用 numa, 或者在vmlinux中禁止. 
net.ipv4.ip_local_port_range = 40000 65535    
# 本地自动分配的TCP, UDP端口号范围
# vm.nr_hugepages = 66536    
建议shared buffer设置超过64GB使用大页,页大小 /proc/meminfo Hugepagesize

配置OS资源限

# vi /etc/security/limits.conf
 
* soft    nofile  1024000
* hard    nofile  1024000
* soft    nproc   unlimited
* hard    nproc   unlimited
* soft    core    unlimited
* hard    core    unlimited
* soft    memlock unlimited
* hard    memlock unlimited

配置OS防火

(建议按业务场景设置,我这里先清掉)

iptables -F

selinux

如果没有这方面的需求,建议禁用

# vi /etc/sysconfig/selinux 
 
SELINUX=disabled
SELINUXTYPE=targeted

部署文件系

注意SSD对齐,延长寿命,避免写放大。

parted -s /dev/sda mklabel gpt
parted -s /dev/sda mkpart primary 1MiB 100%

格式化

mkfs.ext4 /dev/sda1 -m 0 -O extent,uninit_bg -E lazy_itable_init=1 -T largefile -L u01

建议使用的ext4 mount选项

# vi /etc/fstab
 
LABEL=u01 /u01     ext4        defaults,noatime,nodiratime,nodelalloc,barrier=0,data=writeback    0 0
 
 
# mount -a

编译

建议使用较新的编译器,安装 gcc 6.2.0 

cd ~
tar -jxvf gcc6.2.0.tar.bz2
tar -jxvf python2.7.12.tar.bz2
 
 
# vi /etc/ld.so.conf
 
/home/digoal/gcc6.2.0/lib
/home/digoal/gcc6.2.0/lib64
/home/digoal/python2.7.12/lib
 
# ldconfig

环境变量

# vi ~/env_pg.sh
 
export PS1="$USER@`/bin/hostname -s`-> "
export PGPORT=$1
export PGDATA=/$2/digoal/pg_root$PGPORT
export.utf8
export PGHOME=/home/digoal/pgsql9.6
export LD_LIBRARY_PATH=/home/digoal/gcc6.2.0/lib:/home/digoal/gcc6.2.0/lib64:/home/digoal/python2.7.12/lib:$PGHOME/lib:/lib64:/usr/lib64:/usr/local/lib64:/lib:/usr/lib:/usr/local/lib:$LD_LIBRARY_PATH
export PATH=/home/digoal/gcc6.2.0/bin:/home/digoal/python2.7.12/bin:/home/digoal/cmake3.6.3/bin:$PGHOME/bin:$PATH:.
export DATE=`date +"%Y%m%d%H%M"`
export MANPATH=$PGHOME/share/man:$MANPATH
export PGHOST=$PGDATA
export PGUSER=postgres
export PGDATABASE=postgres
alias rm='rm -i'
alias ll='ls -lh'
unalias vi
 

编译PostgreSQL

建议使用NAMED_POSIX_SEMAPHORES

. ~/env_pg.sh 1921 u01
 
cd postgresql-9.6.1
export USE_NAMED_POSIX_SEMAPHORES=1
LIBS=-lpthread CC="/home/digoal/gcc6.2.0/bin/gcc" CFLAGS="-O3 -flto" ./configure --prefix=/home/digoal/pgsql9.6
LIBS=-lpthread CC="/home/digoal/gcc6.2.0/bin/gcc" CFLAGS="-O3 -flto" make world -j 64
LIBS=-lpthread CC="/home/digoal/gcc6.2.0/bin/gcc" CFLAGS="-O3 -flto" make install-world

初始化数据集群

pg_xlog建议放在IOPS最好的分区。

. ~/env_pg.sh 1921 u01
initdb -D $PGDATA -E UTF8 --locale=C -U postgres -X /u02/digoal/pg_xlog$PGPORT

配置postgresql.conf

PostgreSQL 9.6, 512G内存主机为例

最佳到文件末尾即可,重复的会以末尾的作为有效值。  
 
$ vi postgresql.conf
 
listen_addresses = '0.0.0.0'
port = 1921
max_connections = 5000
unix_socket_directories = '.'
tcp_keepalives_idle = 60
tcp_keepalives_interval = 10
tcp_keepalives_count = 10
shared_buffers = 128GB
maintenance_work_mem = 4GB
dynamic_shared_memory_type = posix
vacuum_cost_delay = 0
bgwriter_delay = 10ms
bgwriter_lru_maxpages = 1000
bgwriter_lru_multiplier = 10.0
bgwriter_flush_after = 0
max_parallel_workers_per_gather = 0
old_snapshot_threshold = -1
backend_flush_after = 0
wal_level = replica
synchronous_commit = off
full_page_writes = on
wal_buffers = 1GB
wal_writer_delay = 10ms
wal_writer_flush_after = 0
checkpoint_timeout = 30min
max_wal_size = 256GB
min_wal_size = 64GB
checkpoint_completion_target = 0.05
checkpoint_flush_after = 0
max_wal_senders = 5
random_page_cost = 1.0
parallel_tuple_cost = 0
parallel_setup_cost = 0
min_parallel_relation_size = 0
effective_cache_size = 300GB
force_parallel_mode = off
log_destination = 'csvlog'
logging_collector = on
log_truncate_on_rotation = on
log_checkpoints = on
log_connections = on
log_disconnections = on
log_error_verbosity = verbose
log_timezone = 'PRC'
autovacuum = on
log_autovacuum_min_duration = 0
autovacuum_max_workers = 16
autovacuum_naptime = 15s
autovacuum_vacuum_scale_factor = 0.02
autovacuum_analyze_scale_factor = 0.01
vacuum_freeze_table_age = 1500000000
vacuum_multixact_freeze_table_age = 1500000000
datestyle = 'iso, mdy'
timezone = 'PRC'
lc_messages = 'C'
lc_monetary = 'C'
lc_numeric = 'C'
lc_time = 'C'
default_text_search_config = 'pg_catalog.english'
shared_preload_libraries='pg_stat_statements'

配置pg_hba.conf

避免不必要的访问,开放允许的访问,建议务必使用密码访问。

$ vi pg_hba.conf
 
host all all 0.0.0.0/0 md5

数据

pg_ctl start

好了,你的PostgreSQL数据库基本上部署好了,可以愉快的玩耍了。


文章评论

  • 登录后评论

点击排行