社区生态 | openEuler、龙蜥Anolis、统信UOS三大主流操作系统下编译GreatSQL二进制包

212023.04

社区生态 | openEuler、龙蜥Anolis、统信UOS三大主流操作系统下编译GreatSQL二进制包

2023.04.21

为了更好地支持更多操作系统及相关生态，GreatSQL社区决定发布openEuler、龙蜥Anolis、统信UOS三个操作系统下的GreatSQL二进制包。相应的二进制包可以访问gitee.com上的 GreatSQL项目 https://gitee.com/GreatSQL/GreatSQL/releases/tag/GreatSQL-8.0.25-17下载。

本文简要记录在这三个操作系统下编译GreatSQL二进制包的过程。

1. 编译环境
2. 编译前准备工作
3. 编译GreatSQL
4. 初始化并启动GreatSQL数据库
5. 运行sysbench测试
附录：编译sysbench

1. 编译环境

本次编译都是采用鲲鹏916这个型号的CPU（泰山2280服务器系列）：

    $ lscpu
    Architecture:          aarch64
    Byte Order:            Little Endian
    CPU(s):                64
    On-line CPU(s) list:   0-63
    Thread(s) per core:    1
    Core(s) per socket:    32
    Socket(s):             2
    NUMA node(s):          4
    Model:                 2
    BogoMIPS:              100.00
    L1d cache:             32K
    L1i cache:             48K
    L2 cache:              1024K
    L3 cache:              16384K
    NUMA node0 CPU(s):     0-15
    NUMA node1 CPU(s):     16-31
    NUMA node2 CPU(s):     32-47
    NUMA node3 CPU(s):     48-63
    Flags:                 fp asimd evtstrm aes pmull sha1 sha2 crc32 cpuid

上述 lscpu 是在物理机上执行的，实际编译环境则是在这个物理机上运行的虚机中，分配了8个CPU、16G内存。

查看操作系统发行版本

openEuler

    $ cat /etc/os-release

    NAME="openEuler"
    VERSION="22.03 LTS"
    ID="openEuler"
    VERSION_ID="22.03"
    PRETTY_NAME="openEuler 22.03 LTS"
    ANSI_COLOR="0;31"

龙蜥Anolis

    $ cat /etc/os-release
    NAME="Anolis OS"
    VERSION="8.6"
    ID="anolis"
    ID_LIKE="rhel fedora centos"
    VERSION_ID="8.6"
    PLATFORM_ID="platform:an8"
    PRETTY_NAME="Anolis OS 8.6"
    ANSI_COLOR="0;31"
    HOME_URL="https://openanolis.cn/"

统信UOS

    $ cat /etc/os-release
    PRETTY_NAME="UnionTech OS Server 20"
    NAME="UnionTech OS Server 20"
    VERSION_ID="20"
    VERSION="20"
    ID="uos"
    HOME_URL="https://www.chinauos.com/"
    BUG_REPORT_URL="https://bbs.chinauos.com/"
    VERSION_CODENAME="kongzi"
    PLATFORM_ID="platform:uelc20"
    [root@yejr-uos-aarch64 ~]#
    [root@yejr-uos-aarch64 ~]#
    $ cat /etc/uos-release
    UnionTech OS Server release 20 (kongzi)

并且都采用OS中预设的默认YUM源

    $ cat openEuler.repo
    #generic-repos is licensed under the Mulan PSL v2.
    #You can use this software according to the terms and conditions of the Mulan PSL v2.
    #You may obtain a copy of Mulan PSL v2 at:
    #    http://license.coscl.org.cn/MulanPSL2
    #THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR
    #IMPLIED, INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY OR FIT FOR A PARTICULAR
    #PURPOSE.
    #See the Mulan PSL v2 for more details.

    [OS]
    name=OS
    baseurl=http://repo.openeuler.org/openEuler-22.03-LTS/OS/$basearch/
    enabled=1
    gpgcheck=1
    gpgkey=http://repo.openeuler.org/openEuler-22.03-LTS/OS/$basearch/RPM-GPG-KEY-openEuler

    [everything]
    name=everything
    baseurl=http://repo.openeuler.org/openEuler-22.03-LTS/everything/$basearch/
    enabled=1
    gpgcheck=1
    gpgkey=http://repo.openeuler.org/openEuler-22.03-LTS/everything/$basearch/RPM-GPG-KEY-openEuler
    ...

龙蜥Anolis

    $ cat AnolisOS-AppStream.repo
    [AppStream]
    name=AnolisOS-$releasever - AppStream
    baseurl=http://mirrors.openanolis.cn/anolis/$releasever/AppStream/$basearch/os
    enabled=1
    gpgkey=file:///etc/pki/rpm-gpg/RPM-GPG-KEY-ANOLIS
    gpgcheck=1

统信UOS

    $ cat UniontechOS.repo
    [UniontechOS-$releasever-AppStream]
    name = UniontechOS $releasever AppStream
    baseurl = https://enterprise-c-packages.chinauos.com/server-enterprise-c/kongzi/1050/AppStream/$basearch
    enabled = 1
    username=$auth_u
    password=$auth_p
    gpgkey = file:///etc/pki/rpm-gpg/RPM-GPG-KEY-uos-release
    gpgcheck = 0
    skip_if_unavailable = 1

    [UniontechOS-$releasever-BaseOS]
    name = UniontechOS $releasever BaseOS
    baseurl = https://enterprise-c-packages.chinauos.com/server-enterprise-c/kongzi/1050/BaseOS/$basearch
    enabled = 1
    username=$auth_u
    password=$auth_p
    gpgkey = file:///etc/pki/rpm-gpg/RPM-GPG-KEY-uos-release
    gpgcheck = 0
    skip_if_unavailable = 1
    ...

2. 编译前准备工作

参考文档麒麟OS+龙芯环境编译GreatSQL，提前安装必要的一些基础包

    $ dnf makecache
    $ dnf install --skip-broken -y automake bison bison-devel boost-devel bzip2 bzip2-devel clang \
    cmake cmake3 diffutils expat-devel file flex gcc gcc-c++ git jemalloc jemalloc-devel \
    graphviz libaio-devel libarchive libcurl-devel libevent libevent-devel libverto-libevent libevent-doc libffi-devel libicu-devel libssh \
    libtirpc libtirpc-devel libtool libxml2-devel libzstd libzstd-devel lz4-devel \
    lz4-static make ncurses-devel ncurses-libs net-tools numactl numactl-devel numactl-libs openldap-clients \
    openldap-devel openssl openssl-devel pam pam-devel perl perl-Env perl-JSON perl-Memoize \
    perl-Time-HiRes pkg-config psmisc re2-devel readline-devel \
    snappy-devel tar time unzip vim wget zlib-devel

openEuler下就可以根据安装文档麒麟OS+龙芯环境编译GreatSQL中列出的所有包，包括 jemalloc 包。

下载安装jemalloc rpm包（rpm包依赖glibc版本，可能无法直接使用，可以自行下载源码包编译）

https://fedora.pkgs.org/36/fedora-aarch64/jemalloc-5.2.1-7.fc36.aarch64.rpm.html
https://fedora.pkgs.org/36/fedora-aarch64/jemalloc-devel-5.2.1-7.fc36.aarch64.rpm.html
源码包 https://sourceforge.net/projects/jemalloc.mirror/files/

当然了，jemalloc并库不是必须的，用它的好处是可以优化内存管理性能等。有条件的话尽量启用，实在搞不定就放弃。

如果需要手动编译安装jemalloc，参考下面的方法即可：

    $ tar zxf jemalloc-5.2.1.tar.gz
    $ mv jemalloc-jemalloc-886e40b/
    $ ./autogen.sh
    $ ./configure --prefix=/usr && make && make install

分别下载几个编译过程中需要的依赖包：

boost, https://boostorg.jfrog.io/artifactory/main/release/1.73.0/source/boost_1_73_0.tar.gz
patchelf, https://github.com/NixOS/patchelf/archive/refs/tags/0.14.tar.gz, 下载后重命名为 patchelf-0.14.tar.gz
rpcsvc-proto, https://github.com/thkukuk/rpcsvc-proto/releases/download/v1.4/rpcsvc-proto-1.4.tar.gz

下载GreatSQL源码包：https://product.greatdb.com/GreatSQL-8.0.25-17/greatsql-8.0.25-17.tar.gz

将所有的源码包都放在 /opt 目录下。

编译安装patchelf：

    $ cd /opt && tar zxvf patchelf-0.14.tar.gz && cd patchelf-0.14 && ./bootstrap.sh && ./configure && make && make install

编译安装rpcsvc-proto：

    $ cd /opt && tar zxvf rpcsvc-proto-1.4.tar.gz && cd rpcsvc-proto-1.4/ && ./configure && make && make install

编译安装rpcsvc-proto：

    [root@ky10 ~]# cd /opt && tar zxvf rpcsvc-proto-1.4.tar.gz && cd rpcsvc-proto-1.4/ && ./configure && make && make install

确认glibc版本：

    $ ldd --version

    ldd (GNU libc) 2.28
    Copyright (C) 2018 Free Software Foundation, Inc.
    This is free software; see the source for copying conditions.  There is NO
    warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
    Written by Roland McGrath and Ulrich Drepper.

在本次编译过程中，openEuler 2203的glibc版本是2.34。

3. 编译GreatSQL

接下来编译GreatSQL二进制包

    $ cat /opt/greatsql-build-tarball.sh

    #!/bin/bash
    MAJOR_VERSION=8
    MINOR_VERSION=0
    PATCH_VERSION=25
    RELEASE=17
    REVISION="4733775f703"
    GLIBC=`ldd --version | grep ldd | tail -n 1 | awk '{print $NF}'`
    ARCH=aarch64
    OS=openEuler
    PKG_NAME=GreatSQL-${MAJOR_VERSION}.${MINOR_VERSION}.${PATCH_VERSION}-${RELEASE}-${OS}-glibc${GLIBC}-${ARCH}BASE_DIR=/usr/local/${PKG_NAME}
    SRC_DIR=/opt
    BOOST_SOURCE_DIR=boost_1_73_0
    GREATSQL_SOURCE_DIR=greatsql-8.0.25-17
    JOBS=`nproc`

    # 如果你的OS环境下已安装jemalloc，建议也启用jemalloc编译选项
    # 如果没有安装jemalloc，则将本行参数注释掉
    CMAKE_EXE_LINKER_FLAGS=" -ljemalloc "

    cd ${SRC_DIR}/${GREATSQL_SOURCE_DIR} && \
    rm -fr bld && \
    mkdir bld && \
    cd bld && \
    cmake .. -DBOOST_INCLUDE_DIR=${SRC_DIR}/${BOOST_SOURCE_DIR} \
    -DLOCAL_BOOST_DIR=${SRC_DIR}/${BOOST_SOURCE_DIR} \
    -DCMAKE_INSTALL_PREFIX=${BASE_DIR} -DWITH_ZLIB=bundled \
    -DWITH_NUMA=ON -DCMAKE_EXE_LINKER_FLAGS="${CMAKE_EXE_LINKER_FLAGS}" \
    -DCMAKE_BUILD_TYPE=RelWithDebInfo -DBUILD_CONFIG=mysql_release \
    -DWITH_TOKUDB=OFF -DWITH_ROCKSDB=OFF \
    -DCOMPILATION_COMMENT="GreatSQL (GPL), Release ${RELEASE}, Revision ${REVISION}" \
    -DMAJOR_VERSION=${MAJOR_VERSION} -DMINOR_VERSION=${MINOR_VERSION} -DPATCH_VERSION=${PATCH_VERSION} \
    -DWITH_UNIT_TESTS=OFF -DWITH_NDBCLUSTER=OFF -DWITH_SSL=system -DWITH_SYSTEMD=ON \
    -DWITH_LIBEVENT=system \
    && make -j${JOBS} && make -j${JOBS} install

不出意外的话，就可以编译生成二进制文件了。

4. 初始化并启动GreatSQL数据库

GreatSQL初始化，my.cnf可以参考这份模板：https://gitee.com/GreatSQL/GreatSQL-Doc/blob/master/docs/my.cnf-example-greatsql-8.0.25-17

    $ groupadd mysql && useradd -g mysql mysql -s /sbin/nologin -d /dev/null

    $ echo '/usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/lib/' > /etc/ld.so.conf.d/greatsql.conf

    $ ldconfig -p | grep -i percona
     libperconaserverclient.so.21 (libc6,AArch64) => /usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/lib/libperconaserverclient.so.21
     libperconaserverclient.so (libc6,AArch64) => /usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/lib/libperconaserverclient.so
     
     
    # 确保没有找不到的动态库
    $ ldd /usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/bin/mysqld | grep -i not

    $ ./bin/mysqld --defaults-file=/etc/my.cnf --initialize-insecure

    $ ./bin/mysqld --defaults-file=/etc/my.cnf &

    $ lsof -p `pidof mysqld` | grep -i jemalloc
    mysqld  85204 mysql  mem       REG                8,3     471696   3329101 /usr/lib64/libjemalloc.so.2

5. 运行sysbench测试

准备跑一轮sysbench测试

    #先设置PATH
    $ export PATH=$PATH:/usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/bin

    $ mysqladmin create sbtest

    $ cd /usr/local/share/sysbench/
    $ sysbench /usr/local/share/sysbench/oltp_read_write.lua --mysql-host=localhost --mysql-user=root --mysql-password="" --mysql-socket=/usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/data/mysql.sock --mysql-db=sbtest --db-driver=mysql --tables=10 --table_size=10000 prepare

    $ for i in $(seq 1 3);do sysbench /usr/local/share/sysbench/oltp_read_write.lua --mysql-host=localhost --mysql-user=root --mysql-password="" --mysql-socket=/usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/data/mysql.sock --mysql-db=sbtest --db-driver=mysql --tables=10 --table_size=10000 --report-interval=1 --threads=8 --rand-type=uniform --db-ps-mode=disable  --time=900 run > greatsql-802517-$i.log; sleep 300; done

附录：编译sysbench

    #先做个动态库软链接
    $ cd /usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/lib/
    $ ln -s libperconaserverclient.so libmysqlclient.so
    $ cd /tmp/sysbench/
    $ ./autogen.sh
    $ ./configure --with-mysql-includes=/usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/include/ --with-mysql-libs=/usr/local/GreatSQL-8.0.25-17-openEuler-glibc2.34-aarch64/lib/ && make && make install

222024.11: 软件业用新创造价值
2024.11.22

222024.11: 喜报 | 万里数据库入围四川省2024年基础软件框架协议采购项目
2024.11.22

222024.11: 展会 | 聚焦数智未来万里数据库受邀操作系统大会 & openEuler Summit 2024
2024.11.22