安装分布式数据库集群)
CloudberryDB 一个 Greenplum Database 分布式数据库开源版本的衍生项目针对开源 Greenplum Database 优化的地方 CloudberryDB制定了路线图https://github.com/orgs/cloudberrydb/discussions/369并在逐步改进 。GitHub 地址https://github.com/cloudberrydb/cloudberrydb官方文档产品特性 | Apache Cloudberry (Incubating)Cloudberry Database 基于最新的 PostgreSQL 14.4 内核是当前最先进的成熟开源 MPP 数据库之一具备高并发、高可用等多种特性可以对复杂任务进行快速高效计算以满足海量数据管理和计算的需求目前在多个领域都有着广泛应用。https://cloudberry.apache.org/docs/准备部署的3台服务器信息如下[rootdfq-cloudberrydb-01 ~]# cat /etc/os-release NAMERocky Linux VERSION8.6 (Green Obsidian) IDrocky ID_LIKErhel centos fedora VERSION_ID8.6 PLATFORM_IDplatform:el8 PRETTY_NAMERocky Linux 8.6 (Green Obsidian) ANSI_COLOR0;32 CPE_NAMEcpe:/o:rocky:rocky:8:GA HOME_URLhttps://rockylinux.org/ BUG_REPORT_URLhttps://bugs.rockylinux.org/ ROCKY_SUPPORT_PRODUCTRocky Linux ROCKY_SUPPORT_PRODUCT_VERSION8 REDHAT_SUPPORT_PRODUCTRocky Linux REDHAT_SUPPORT_PRODUCT_VERSION8通过源码编译CloudberryDB下载CloudberryDB源码git clone https://github.com/cloudberrydb/cloudberrydb.git cd cloudberrydb git checkout 1.5.3安装依赖项sudo dnf groupinstall -y Development Tools sudo dnf install -y epel-release sudo dnf install -y apr-devel bison bzip2-devel cmake3 flex gcc gcc-c krb5-devel libcurl-devel libevent-devel libkadm5 libxml2-devel libzstd-devel openssl-devel perl-ExtUtils-Embed python3-devel python3-pip readline-devel xerces-c-devel zlib-devel ~/cloudberrydb/deploy/build/README.Rhel-Rocky.bash pip3.9 install -i https://pypi.tuna.tsinghua.edu.cn/simple -r ~/cloudberrydb/python-dependencies.txt编译CloudberryDB设置安装路径为/usr/local/cloudberrydbcd cloudberrydb ./configure --with-perl --with-python --with-libxml --with-gssapi --prefix/usr/local/cloudberrydb make -j8 make install部署前的准备下面的步骤需要在三台服务器都一次执行一次使用root用户来执行命令解压编译好的CloudberryDB安装包到/usr/local目录下然后使用下面的命令分别修改每台服务器的主机名hostnamectl set-hostname dfq-cloudberrydb-0x禁用SELINUX和防火墙setenforce 0 systemctl stop firewalld systemctl disable firewalld修改/etc/hosts文件包含CloudberryDB所有主机别名与其网络IP地址的映射10.197.165.181 dfq-cloudberrydb-01 10.197.165.182 dfq-cloudberrydb-02 10.197.165.183 dfq-cloudberrydb-03配置内核参数在/etc/sysctl.conf文件中添加下面的参数# kernel.shmall _PHYS_PAGES / 2 kernel.shmall 197951838 # kernel.shmmax kernel.shmall * PAGE_SIZE kernel.shmmax 810810728448 kernel.shmmni 4096 vm.overcommit_memory 2 vm.overcommit_ratio 95 net.ipv4.ip_local_port_range 10000 65535 kernel.sem 250 2048000 200 8192 kernel.sysrq 1 kernel.core_uses_pid 1 kernel.msgmnb 65536 kernel.msgmax 65536 kernel.msgmni 2048 net.ipv4.tcp_syncookies 1 net.ipv4.conf.default.accept_source_route 0 net.ipv4.tcp_max_syn_backlog 4096 net.ipv4.conf.all.arp_filter 1 net.ipv4.ipfrag_high_thresh 41943040 net.ipv4.ipfrag_low_thresh 31457280 net.ipv4.ipfrag_time 60 net.core.netdev_max_backlog 10000 net.core.rmem_max 2097152 net.core.wmem_max 2097152 vm.swappiness 10 vm.zone_reclaim_mode 0 vm.dirty_expire_centisecs 500 vm.dirty_writeback_centisecs 100 vm.dirty_background_ratio 0 vm.dirty_ratio 0 vm.dirty_background_bytes 1610612736 vm.dirty_bytes 4294967296 kernel.core_pattern/var/core/core.%h.%t执行下面的命令使参数生效sysctl -p资源限制设置在/etc/security/limits.conf文件中添加下面的参数对软硬件资源用量进行限制*soft nofile 524288 *hard nofile 524288 *soft nproc 131072 *hard nproc 131072 * soft core unlimited创建gpadmin管理用户groupadd -g 520 gpadmin useradd -g 520 -u 520 -m -d /home/gpadmin/ -s /bin/bash gpadmin passwd gpadmin编译好的CloudberryDB软件包在/usr/local/cloudberrydb路径下为gpadmin用户授予安装目录的权限chown -R gpadmin:gpadmin /usr/local/cloudberrydb在所有的节点安装依赖sudo dnf install -y apr-devel bison bzip2-devel cmake3 epel-release flex gcc gcc-c krb5-devel libcurl-devel libevent-devel libkadm5 libxml2-devel libzstd-devel openssl-devel python39 python39-devel python39-psutil python3-pip perl-ExtUtils-MakeMaker.noarch perl-ExtUtils-Embed.noarch readline-devel rsync xerces-c-devel zlib-devel python3-psutil python3-pyyaml python3-psycopg2 sudo yum --enablerepopowertools install -y libyaml-devel su - gpadmin pip3.9 install -i https://pypi.tuna.tsinghua.edu.cn/simple psutil5.7.0 pygresql5.2 pyyaml5.3.1 sudo rm -rf /usr/bin/python3 sudo ln -s /usr/bin/python3.9 /usr/bin/python3配置节点间的互信切换到gpadmin用户后续的操作均使用gpadmin用户进行su - gpadmin创建节点配置信息在/home/gpadmin/all_hosts里填入下面的信息dfq-cloudberrydb-01 dfq-cloudberrydb-02 dfq-cloudberrydb-03在/home/gpadmin/seg_hosts里填入segment节点的信息dfq-cloudberrydb-02 dfq-cloudberrydb-03在每台主机上执行下面的命令生成SSH密钥ssh-keygen在每台主机上执行下面的命令来配置免密ssh-copy-id dfq-cloudberrydb-01 ssh-copy-id dfq-cloudberrydb-02 ssh-copy-id dfq-cloudberrydb-03初始化CloudberryDB在Coordinator节点使用gpadmin用户来操作在所有的节点执行在~/.bashrc文件中添加下面的命令source /usr/local/cloudberrydb/greenplum_path.sh执行下面的命令使命令生效source ~/.bashrc在Coordinator节点使用下面的命令来为Segment节点创建数据目录和mirror目录gpssh -f seg_hosts mkdir -p /data/primary mkdir -p /data/mirror在Coordinator节点上创建数据目录mkdir -p /data/coordinator在Coordinator节点的主机上往~/.bashrc文件中添加下面的命令export COORDINATOR_DATA_DIRECTORY/data/coordinator/gpseg-1执行下面的命令使文件生效source ~/.bashrc配置启动脚本在Coordinator节点所在主机上复制模板文件cp $GPHOME/docs/cli_help/gpconfigs/gpinitsystem_config .修改gpinitsystem_config文件端口、Coordinator节点、Segment节点、Mirror节点的信息配置成之前配置的将DATA_DIRECTORY修改为Segment计算节点的数据目录/data/primary将COORDINATOR_HOSTNAME修改为dfq-cloudberrydb-01将COORDINATOR_DIRECTORY修改为Coordinator节点数据目录/data/coordinator将MIRROR_DATA_DIRECTORY修改为Mirror的数据目录/data/mirror初始化CloudberryDB数据库在gpadmin用户下使用下面的命令初始化CloudberryDBgpinitsystem -c gpinitsystem_config -h /home/gpadmin/seg_hosts初始化完成后将下面的环境变量配置到~/.bashrc里export PGDATABASEpostgres export PGHOST127.0.0.1 export PGUSERgpadmin export PGPORT5432重新加载~/.bashrcsource ~/.bashrc使用psql登录数据库[gpadmindfq-cloudberrydb-01 ~]$ psql psql (14.4, server 14.4) Type help for help. postgres# SELECT version(); version ----------------------------------------------------------------------------------- PostgreSQL 14.4 (Cloudberry Database 1.5.3dev.57.ge1d276f80e build dev) on x86_64-pc-li nux-gnu, compiled by gcc (GCC) 8.5.0 20210514 (Red Hat 8.5.0-22), 64-bit compiled on Jun 19 2024 09:20:42 (1 row)【Docker中快速部署】1 docker pull registry.cn-hangzhou.aliyuncs.com/lhrbest/cbdb:1.5.4 2 docker tag registry.cn-hangzhou.aliyuncs.com/lhrbest/cbdb:1.5.4 lhrbest/cbdb:1.5.4 3 4 5 docker rm -f cbdb 6 docker run -d --name cbdb -h cbdb \ 7 -p 24432:5432 -p 2422:22 \ 8 -v /sys/fs/cgroup:/sys/fs/cgroup \ 9 --privilegedtrue lhrbest/cbdb:1.5.4 \ 10 /usr/sbin/init 11 12 13 docker exec -it cbdb bash 14 15 16 17 su - gpadmin 18 gpstart -a【自带的维护工具】\$GPHOME/bin/ 目录下需要 source $GPHOME/greenplum_path.shgpssh 并行执行 shell 命令gpscp 并行文件复制类似 scpgpssh-exkeys 在集群节点间交换 SSH 公钥建立免密登录gpconfig 修改 GUC 参数配置