一、编译环境

  • Linux version 3.10.0-327.el7.x86_64 (builder@kbuilder.dev.centos.org) (gcc version 4.8.3 20140911 (Red Hat 4.8.3-9) (GCC) ) 
  • gpdb-5.16.0

二、编译过程

2.1、下载greenplum源码

可从该位置下载需要的版本:https://github.com/greenplum-db/gpdb/releases/(此处使用gpdb-5.16.0)

2.2、解压

下载后再centos中解压到指定目录,如/home/gpdb_src

2.3、安装必备的工具

  • psutil
  • lockfile (>= 0.9.1)
  • paramiko
  • setuptools
  • conan

注意,以上工具使用pip(>=7.x.x)安装,如果没有pip,可以参考https://pip.pypa.io/en/stable/installing/步骤进行安装。

如果pip版本太低,可使用如下命令升级:

pip install --upgrade pip

2.4、配置ld.so.conf

由于我们使用的是CentOS系统, 所以一定要加上/usr/local/lib和/usr/local/lib64的路径到/etc/ld.so.conf文件中,这一步尤为重要,目的是将常用的动态函数库加载到内存中,可以提高函数库的访问效率。另外,如果不配置这一步,在后面编译的过程中会遇到一些奇奇怪怪的缺少引用的错误

修改后的文件内容类似如下:

centos7部署单机zk kafka centos7 kdump_python

修改完成后保存,之后运行ldconfig命令

2.5、创建gpadmin用户和配置ssh

这一步可以使用源码中自带的脚本完成,脚本位置如下:/home/gpdb_src/gpdb-5.16.0/concourse/scripts/setup_gpadmin_user.bash,执行如下命令:

cd /home
bash ./gpdb_src/concourse/scripts/setup_gpadmin_user.bash

这里说明下,这一步如果使用脚本操作,一定要把源码文件放到gpdb_src中,执行脚本时需要进入到与gpdb_src同级的目录执行,如果要手动配置也是可以的。具体的操作也不复杂,有兴趣的朋友可以读下setup_gpadmin_user.bash脚本。

执行完成后,进行如下两步操作验证:

centos7部署单机zk kafka centos7 kdump_python_02

如果切换账户正常并且ssh正常,不需要输入密码,这说明这一步成功

2.6、修改内核参数

sudo bash -c 'cat >> /etc/sysctl.conf <<-EOF
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

EOF'

sudo bash -c 'cat >> /etc/security/limits.conf <<-EOF
* soft nofile 65536
* hard nofile 65536
* soft nproc 131072
* hard nproc 131072

EOF'

sudo bash -c 'cat >> /etc/ld.so.conf <<-EOF
/usr/local/lib

EOF'

这一步使用root账户操作,并且很重要,否则在后期使用greenplum的时候会遇到问题,详细的问题后面会解析。

2.7、编译GPORCA(可选)

这里提一下,如果不需要优化器这一步也可以跳过。毕竟greenplum在不安装orca的情况下使用的是legacy优化器。

GPORCA是postgresql新一代的优化器,在性能上有很大提升,至于为什么要有两个优化器,其实是有历史原因的,早期legacy是针对单节点PostgreSQL而构建的,主要应用于OLTP场景,现在的greenplum使用MPP,主要应用场景变为OLAP场景,legacy对此虽然进行了修改,但是从架构设计上,使得其维护和添加新的功能越来越困难,所以有了GPORCA优化器。

首先,编译gporca前需要安装如下必备工具:

  • CMake(>=3.1)
  • xerces-c(>=3.1.x)
  • re2c(>= 0.11.3)

2.7.1、自动编译

cd /home/gpdb_src/gpdb-5.16.0/depends
./configure
make
make install_local

 

2.7.2 手工编译

特别注意,安装前一定要确认优化器当前的版本是否和greenplum版本相匹配,如果不匹配将无法编译greenplum,详情参见问题6

1) 下载安装cmake

进入https://cmake.org/download/页面下载自己需要的版本

2) 下载安装gp-xerces

可以使用git下载:https://github.com/greenplum-db/gp-xerces.git

3)  下载安装re2c

进入http://re2c.org/install/install.html页面下载自己需要的版本

安装re3c是由于配置ninja时需要

4) 下载安装Ninja

可以使用git下载:https://github.com/ninja-build/ninja.git

下载后进入ninja目录执行如下命令:

./configure.py --bootstrap

5) 下载安装gporca

可以使用git下载:https://github.com/greenplum-db/gporca.git

解压后进入目录执行如下命令安装:

cmake -GNinja -H. -Bbuild
ninja install -C build

待安装完成后,进入/gporca/build目录,执行ctest命令进行检查

如果最后输出类似如下结果:

100% tests passed, 0 tests failed out of 119

Total Test time (real) = 195.48 sec

这说明编译成功了。

 2.8 编译greenplum

执行如下命令:

cd /home/gpdb_src/gpdb-5.16.0/gpdb-5.16.0
./configure --with-perl --with-python --with-libxml --with-gssapi --prefix=/usr/local/gpdb
make -j8      #代替make命令,并行的编译,提高编译速度

注意:这里如果跳过了第七步(不使用GPORCA优化器),则configure时需要使用如下命令:

./configure --with-perl --with-python --with-libxml --with-gssapi --disable-orca --prefix=/usr/local/gpdb

configure有很多的组件可以加上,如果需要添加自定义可以使用./configure --help查看,加上自己需要的组件即可

 通过以上步骤,就完成了编译。虽然本人已经力求写的详细,但是还是略过了一些自认为大家一看到就能知道原因,同时环境的差异也会导致一些本人没有遇到过的错误,所以有理由相信大家编译的时候还是会遇到一些问题,如果遇到了本文没有涉及到的错误可以告诉我,大家一起讨论。

安装greenplum

cd /home/gpdb_src/gpdb-5.16.0
make -j8 install  #安装

创建测试集群

如下命令需要使用之前创建的数据库管理员账户运行,否则会报如下错误:

centos7部署单机zk kafka centos7 kdump_centos7部署单机zk kafka_03

 

su - gpadmin
source /usr/local/gpdb/greenplum_path.sh #将greenplum的运行信息加入到shell环境中
cd /home/gpdb_src/gpdb-5.16.0/gpAux/gpdemo
source gpdemo-env.sh
make create-demo-cluster

 

centos7部署单机zk kafka centos7 kdump_python_04

可以使用如下命令处理:

yum remove python-gssapi.x86_64

 

三、错误处理

1)

安装必备工具或者升级pip时可能报类似如下错误:

Could not fetch URL https://pypi.python.org/simple/conan/: There was a problem confirming the ssl certificate: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:765) - skipping

可以执行如下命令:

pip install --trusted-host pypi.org --trusted-host files.pythonhosted.org XXXX
pip install --trusted-host pypi.org --trusted-host files.pythonhosted.org --upgrade pip
XXXX为你要安装的工具

如果以上都不解决问题,可是使用-i 参数指定国内的镜像来进行安装

如:

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple psutil
出现如下错误则要修正系统时间:

centos7部署单机zk kafka centos7 kdump_xml_05

当你没有pip,在2.3步安装pip本身都报类似如下错误:
Could not fetch URL https://pypi.python.org/simple/pip/: There was a problem confirming the ssl certificate: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:765) - skipping
  Could not find a version that satisfies the requirement pip (from versions: )
No matching distribution found for pip

可以使用如下命令安装:

sudo yum -y install epel-release
sudo yum -y install python-pip

yum也不好使的情况下,可以尝试如下命令:

python get-pip.py --trusted-host pypi.org --trusted-host

 

2)

checking for xsltCleanupGlobals in -lxslt... no
configure: error: library 'xslt' is required for XSLT support

执行如下命令:yum install libxslt libxslt-devel

3)

/bin/ld: libpq/SUBSYS.o: undefined reference to symbol 'gss_delete_sec_context@@gssapi_krb5_2_MIT'
/bin/ld: note: 'gss_delete_sec_context@@gssapi_krb5_2_MIT' is defined in DSO /lib64/libgssapi_krb5.so.2 so try adding it to the linker command line
/lib64/libgssapi_krb5.so.2: could not read symbols: Invalid operation
collect2: error: ld returned 1 exit status
make[2]: *** [postgres] Error 1
make[2]: Leaving directory `/home/greenplum/src/backend'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/greenplum/src'
make: *** [all] Error 2

类似这种错误基本都是安装步骤中第4步没有配置导致的

4)

checking for flags to link embedded Perl... Can't locate ExtUtils/Embed.pm in @INC (@INC contains: /usr/local/lib64/perl5 /usr/local/share/perl5 /usr/lib64/perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl5 .).
BEGIN failed--compilation aborted.
no
configure: error: could not determine flags for linking embedded Perl.
This probably means that ExtUtils::Embed or ExtUtils::MakeMaker is not installed.

yum install perl-ExtUtils-Embed -y

5)

checking for xmlSaveToBuffer in -lxml2... no
configure: error: library 'xml2' (version >= 2.6.23) is required for XML support

yum install libxml2 libxml2-devel -y

6)

checking Checking ORCA version... configure: error: Your ORCA version is expected to be 2.51.XXX

这说明你装的orca的版本是不匹配的,需要安装匹配版本,那么如何在安装前知道需要什么版本,执行如下命令:

cd /home/gpdb_src/depends
cat conanfile_orca.txt

这里在文件中可以看到如下记录:

[requires]
orca/v2.51.0@gpdb/stable

到github上找到对应版本下载安装即可。

当需要更换版本时,需要清理掉之前已经安装的版本。默认情况下安装在/usr/local/目录下,根据安装的路径,执行如下命令删除已经安装的版本

rm -rf build/*
rm -rf /usr/local/include/naucrates
rm -rf /usr/local/include/gpdbcost
rm -rf /usr/local/include/gpopt
rm -rf /usr/local/include/gpos
rm -rf /usr/local/lib/libnaucrates.so*
rm -rf /usr/local/lib/libgpdbcost.so*
rm -rf /usr/local/lib/libgpopt.so*
rm -rf /usr/local/lib/libgpos.so*

注意:

1、build位于gporca编译目录,其是编译时cmake产生的一些文件

2、如果重新安装了版本一定要执行ldconfig命令,否则新的版本不会在缓存中更新,会一直报上述错误

7)

unable to execute gcc: No such file or directory
    error: command 'gcc' failed with exit status 1

    ----------------------------------------
Command "/usr/bin/python2 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-4lu7Y0/psutil/setup.py';f=getattr(tokenize, 'open', open)(__file__);code=f.read().replace('\r\n', '\n');f.close();exec(compile(code, __file__, 'exec'))" install --record /tmp/pip-YPgXFP-record/install-record.txt --single-version-externally-managed --compile" failed with error code 1 in /tmp/pip-build-4lu7Y0/psutil/

yum install -y gcc

8)如果使用编译参数:--with-gssapi可能会报如下错误:

checking for library containing gss_init_sec_context... no
configure: error: could not find function 'gss_init_sec_context' required for GSSAPI

安装gssapi即可:pip install gssapi

而安装gssapi可能报如下错误:

Complete output from command python setup.py egg_info:
    In distributed package, building from C files...
    Traceback (most recent call last):
      File "<string>", line 1, in <module>
      File "/tmp/pip-build-CHeKis/gssapi/setup.py", line 98, in <module>
        raise Exception("Could not find main GSSAPI shared library.  Please "
    Exception: Could not find main GSSAPI shared library.  Please try setting GSSAPI_MAIN_LIB yourself or setting ENABLE_SUPPORT_DETECTION to 'false'

    ----------------------------------------
Command "python setup.py egg_info" failed with error code 1 in /tmp/pip-build-CHeKis/gssapi/

需要安装如下工具:yum install -y krb5-devel.x86_64

9)

checking for apr-1-config... no
configure: error: apr-1-config is required for gpfdist, unable to find binary

yum install -y apr-devel

10)

checking for library containing event_add... no
configure: error: libevent is required for gpfdist

yum install -y libevent-devel

11)

checking for curl-config... no
*** The curl-config script could not be found. Make sure it is
*** in your path, and that curl is properly installed.
*** Or see http://curl.haxx.se/
configure: error: Library requirements (curl) not met.

yum install -y libcurl-devel.x86_64

12)

checking for bzlib.h... no
configure: error: header file <bzlib.h> is required for bzip2 support

yum install -y bzip2-devel

13)安装psutil时可能报如下错误:

psutil/_psutil_common.c:9:20: fatal error: Python.h: No such file or directory
     #include <Python.h>
                        ^
    compilation terminated.
    error: command 'gcc' failed with exit status 1

    ----------------------------------------
Command "/usr/bin/python2 -u -c "import setuptools, tokenize;__file__='/tmp/pip-build-3gzCd8/psutil/setup.py';exec(compile(getattr(tokenize, 'open', open)(__file__).read().replace('\r\n', '\n'), __file__, 'exec'))" install --record /tmp/pip-EAY9ck-record/install-record.txt --single-version-externally-managed --compile" failed with error code 1 in /tmp/pip-build-3gzCd8/psutil/

yum install -y python-devel.x86_64

 

14)

ERROR: orca/v3.21.0@gpdb/stable: Error in build() method, line 48
        cmake.configure(defs=cmake_defines)
        ConanException: Error 256 while executing cd '/home/gpdb_src/gpdb-5.16.0/depends/.conan/data/orca/v3.21.0/gpdb/stable/build/4f7d6d5032b1a188f98e0c149ef6bf91e76af63e' && cmake -G "Unix Makefiles" -DCMAKE_BUILD_TYPE="Release" -DCONAN_EXPORTED="1" -DCONAN_IN_LOCAL_CACHE="ON" -DCONAN_COMPILER="gcc" -DCONAN_COMPILER_VERSION="4.8" -DCONAN_CXX_FLAGS="-m64" -DCONAN_SHARED_LINKER_FLAGS="-m64" -DCONAN_C_FLAGS="-m64" -DCONAN_LIBCXX="libstdc++" -DBUILD_SHARED_LIBS="ON" -DCMAKE_INSTALL_PREFIX="/home/gpdb_src/gpdb-5.16.0/depends/.conan/data/orca/v3.21.0/gpdb/stable/package/4f7d6d5032b1a188f98e0c149ef6bf91e76af63e" -DCMAKE_INSTALL_BINDIR="bin" -DCMAKE_INSTALL_SBINDIR="bin" -DCMAKE_INSTALL_LIBEXECDIR="bin" -DCMAKE_INSTALL_LIBDIR="lib" -DCMAKE_INSTALL_INCLUDEDIR="include" -DCMAKE_INSTALL_OLDINCLUDEDIR="include" -DCMAKE_INSTALL_DATAROOTDIR="share" -DCMAKE_EXPORT_NO_PACKAGE_REGISTRY="ON" -Wno-dev '/home/gpdb_src/gpdb-5.16.0/depends/.conan/data/orca/v3.21.0/gpdb/stable/build/4f7d6d5032b1a188f98e0c149ef6bf91e76af63e'
make: *** [orca] Error 1

yum install gcc-c++

 

15)该问题可能出现在make阶段

***
ERROR: `bison' is missing on your system. It is needed to create the
file `gram.c'. You can either get bison from a GNU mirror site
or download an official distribution of PostgreSQL, which contains
pre-packaged bison output.
***
make[3]: *** [gram.c] Error 1
make[3]: Leaving directory `/home/gpdb_src/gpdb-5.16.0/src/backend/parser'
make[2]: *** [parser/gram.h] Error 2
make[2]: *** Waiting for unfinished jobs....
AWK='gawk' /bin/sh Gen_fmgrtab.sh pg_proc_combined.h.tmp
make[3]: Leaving directory `/home/gpdb_src/gpdb-5.16.0/src/backend/utils'
make[2]: Leaving directory `/home/gpdb_src/gpdb-5.16.0/src/backend'
make[1]: *** [all] Error 2
make[1]: Leaving directory `/home/gpdb_src/gpdb-5.16.0/src'
make: *** [all] Error 2

同样的flex也可能出现类似问题,解决稍微麻烦一点

yum install -y bison.x86_64 bison-devel.x86_64

 

之后需要make clean

重新执行安装流程

 

15)

In file included from include/reader.h:4:0,
                 from include/gpreader.h:4,
                 from src/gpreader.cpp:1:
include/s3common_headers.h:8:26: fatal error: openssl/hmac.h: No such file or directory
 #include <openssl/hmac.h>
                          ^
compilation terminated.

yum install -y openssl-devel

 

四、参考资料

https://github.com/greenplum-db/gpdb/tree/5.16.0