案例：华为FusionInsight_HD 低配版（三台）安装实例 -- 手动安装

Posted 2022-11-30 袁义锐

本文为安装华为FusionInsight_HD 低配版（三台）安装实例，给小伙伴们参考；

本文选择三台机器安装，实际安装中发现，不能使用华为给的安装引导文件的（FusionInsight V100R002C70SPC200 配置规划工具 ）

如下几点非常重要，供参考

1. 华为安装指导文档要求机器磁盘和内存非常大，但是低配置机器并不影响安装；

2. 消耗内存非常大，每台机器内存基本消耗在20G左右；

3. 实际在安装中并不需要按照华为要求使用oracle jdk；openjdk 一样好用，前提是必须将相应的依赖包补齐；

4. 华为软件需要的依赖包非常多，请执行安装前的检查，本文第5步，可以提前执行；

5. 进入web端后会自动化安装，请注意保持服务器密码一致；

6. redis集群无法通过Web端实现安装；

 

• 安装：

1.上传华为安装文件到主节点/opt目录

[root@SHB-L0120794 redis-cluster]# cd /opt
[root@SHB-L0120794 opt]# ll
total 4543916
lrwxrwxrwx  1 root root          16 Oct 24  2018 chef -> /opt/cloud/chef/
d---------  4 root root        4096 Mar  2 14:30 client
drwxr-xr-x 11 root root         140 Feb 24 11:16 cloud
-rw-r--r--  1 root root   642730898 Feb 24 13:01 FusionInsight_Elk_V100R002C70SPC200_RHEL.tar.gz
-rw-r--r--  1 root root   357649152 Feb 24 13:01 FusionInsight_Flink_V100R002C70SPC200_RHEL.tar.gz
-rw-r--r--  1 root root   949994384 Feb 24 13:01 FusionInsight_HD_V100R002C70SPC200_RHEL.tar.gz
drwxr-xr-x  3 omm  wheel         22 Feb 24 13:01 FusionInsight_Manager
-rw-r--r--  1 root root   615185011 Feb 24 13:01 FusionInsight_Manager_V100R002C70SPC200_RHEL.tar.gz
-rw-r--r--  1 root root  1288870719 Feb 24 13:01 FusionInsight_Porter_V100R002C70SPC200_RHEL.tar.gz
drwxr-xr-x  9 omm  wheel        188 Feb 24 13:01 FusionInsight_SetupTool
-rw-r--r--  1 root root      267594 Feb 24 13:01 FusionInsight_SetupTool_V100R002C70SPC200.tar.gz
-rw-r--r--  1 root root   540351971 Feb 24 13:01 FusionInsight_Spark2x_V100R002C70SPC200_RHEL.tar.gz
-rw-r--r--  1 root root   224148134 Feb 24 13:01 FusionInsight_Streaming_V100R002C70SPC200_RHEL.tar.gz

 

2.解压FusionInsight_Manager文件：

#解压
tar -zxvf FusionInsight_Manager_V100R002C70SPC200_RHEL.tar.gz

 

3.复制：将其他的所有.tar.gz包复制到FusionInsight_Manager文件夹下面：FusionInsight_SetupTool_V100R002C70SPC200.tar.gz除外

#复制
cp FusionInsight_HD_V100R002C70SPC200_RHEL.tar.gz /opt/FusionInsight_Manager/software/packs/

 

4.上传FusionInsight_SetupTool_V100R002C70SPC200.tar.gz到其他两个节点；解压完成后执行如下命令

#复制
cp -r /opt/FusionInsight_SetupTool/preset /opt/preset
#启动
sh /opt/preset/preset.sh

主节点需要多执行如下两行命令：

#授权
chown omm: /opt/FusionInsight_Manager/ -R
chown omm: /opt/FusionInsight_SetupTool -R

 

5. 安装依赖包  所有节点都要执行 执行如下命令行：非常重要

yum -y install ipmitool.x86_64 glibc.i686 json-c.x86_64 libestr.x86_64 libitm.x86_64 lsscsi.x86_64 m2crypto.x86_64 net-snmp-utils.x86_64 nscd.x86_64 nss_compat_ossl.x86_64 OpenIPMI.x86_64 openldap-clients.x86_64 openldap-servers.x86_64 procmail.x86_64 python-dmidecode.x86_64 python-gudev.x86_64 sdparm.x86_64 libitm lsscsi  m2crypto net-snmp-utils nscd nss_compat_ossl OpenIPMI procmail python-dmidecode python-gudev python-simplejson sdparm acpid.x86_64 at.x86_64 bc.x86_64 c-ares.x86_64 checkpolicy.x86_64 crash.x86_64 cups-client.x86_64 cvs.x86_64 cyrus-sasl-plain.x86_64 dos2unix.x86_64 ed.x86_64 gamin.x86_64 hdparm.x86_64 hesiod.x86_64 iptables-services.x86_64 ipvsadm.x86_64 libdhash.x86_64 libipa_hbac.x86_64 libldb.x86_64 libnl.x86_64 libsmbclient.x86_64 libsss_idmap.x86_64 libsss_autofs.x86_64 libsss_sudo.x86_64 libsss_nss_idmap.x86_64 libtar.x86_64 libtdb.x86_64 libusb.x86_64 libwbclient.x86_64 m4.x86_64 mlocate.x86_64 pax.x86_64 pm-utils.x86_64 portreserve.x86_64 pyOpenSSL.x86_64 pytalloc.x86_64 python-dateutil.noarch python-ethtool.x86_64 python-sssdconfig.noarch pytz.noarch redhat-lsb-core.x86_64 redhat-lsb-submod-security.x86_64 rng-tools.x86_64 samba-client-libs.x86_64 samba-common.noarch samba-common-libs.x86_64 samba-common-tools.x86_64 samba-libs.x86_64 spax.x86_64 sssd.x86_64 sssd-ad.x86_64 sssd-client.x86_64 sssd-common.x86_64 sssd-common-pac.x86_64 sssd-ipa.x86_64 sssd-krb5.x86_64 sssd-krb5-common.x86_64 sssd-ldap.x86_64 sssd-proxy.x86_64 system-config-firewall-base.noarch time.x86_64 traceroute.x86_64 tzdata-java.noarch usermode.x86_64 libitm lsscsi  m2crypto net-snmp-utils nscd nss_compat_ossl OpenIPMI procmail python-dmidecode python-gudev  tcl-devel  expect attr.x86_64 bridge-utils.x86_64 cyrus-sasl.x86_64 gdbm-devel.x86_64 jbigkit-libs.x86_64 libcgroup.x86_64 libdb-devel.x86_64 libreport-filesystem.x86_64 libtiff.x86_64 mdadm.x86_64 net-snmp.x86_64 net-snmp-agent-libs.x86_64 numactl.x86_64 numad.x86_64 oddjob.x86_64 oddjob-mkhomedir.x86_64 pam_krb5.x86_64 perl-Compress-Raw-Bzip2.x86_64 perl-Compress-Raw-Zlib.x86_64 perl-DBI.x86_64 perl-ExtUtils-Install.noarch perl-ExtUtils-MakeMaker.noarch perl-ExtUtils-Manifest.noarch perl-ExtUtils-ParseXS.noarch perl-IO-Compress.noarch perl-Net-Daemon.noarch perl-PlRPC.noarch perl-Test-Harness.noarch perl-devel.x86_64 perl-version.x86_64 pyparsing.noarch smartmontools.x86_64 systemtap-sdt-devel.x86_64

#华为整理的需要添加的依赖包位置如下，供参考；使用如下地址中的依赖包，系统会被升级，问题很严重！！！

#华为整理的需要添加的依赖包位置
FusionInsight_SetupTool_V100R002C70SPC200.tar\\FusionInsight_SetupTool_V100R002C70SPC200\\FusionInsight_SetupTool\\preinstall\\script\\function\\add_packages\\conf

也可以通过precheck命令检查，补全依赖包

#进入目录
cd /opt/FusionInsight_SetupTool
#执行依赖包检查命令
./setuptool.sh precheck

其他依赖包（所有节点都要安装）：

1.ldap:

#安装
yum -y install openldap openldap-servers openldap-clients

cp /usr/share/openldap-servers/DB_CONFIG.example /var/lib/ldap/DB_CONFIG

chown ldap:ldap /var/lib/ldap/DB_CONFIG

#启动
systemctl start slapd 

systemctl enable slapd

#查看状态
systemctl status slapd

2. tcl-devel和expect

yum -y install tcl-devel  expect

3.ipmitool

#安装：
yum install ipmitool -y
 
#检查：
rpm -qa |grep ipmi
 
#检查模块：
lsmod  | grep ipmi
 
 
#加载模块：
 
modprobe ipmi_watchdog
modprobe ipmi_poweroff
modprobe ipmi_devintf
modprobe ipmi_si   
#报错：FATAL: Module ipmi_si not found.
 
modprobe ipmi_msghandler
#报错：FATAL: Module ipmi_msghandler not found.
 

 

 

 

6.安装 执行如下安装命令： 一共有6步，请耐心等待； 中间如果遇到报错，请按提示信息放心大胆卸载；

sh /opt/FusionInsight_Manager/software/install.sh -m single

第6步完成之后，会有一个登陆首页的跳转地址：http://主机ip:8080/web/

 

7.设置Internet Explorer浏览器中的： Internet选项  》》 高级 》》取消勾选：“使用TLS1.0”

    该步骤非常重要，不然首页无法进入！！！

 

8.登陆：用户admin,默认密码Admin@123；首次登陆需要修改密码

 

9. 选择默认安装：手工安装，模板不合适

 

10.为避免问题过多，首次安装，切勿贪多；请使用默认DBService和Metadata两个组件安装；

     其他组件待安装完成后，以添加服务的形式安装；

 

因网络问题，无法上传截图；请参考其他安装文档：

https://blog.csdn.net/weixin_37987487/article/details/81632176

 

 

• 报错问题解决：

问题1. 集群部署配置系统环境失败，其他数据节点全部报错：Operation failure for Nodeagent# XX.XX.XX.XX@CD-BigDataDNX.

解决方案： 执行如下命令删除所有除管理节点外其他数据节点的omm用户以及目录

pkill -u omm

userdel -f omm

rm -rf /home/omm

 

问题2.  Ldapclient安装失败

报错信息：

[2020-03-12 11:58:13]RoleInstance initialization failure [ScriptExecutionResult=ScriptExecutionResult [exitCode=1, output=, errMsg=]] for SlapdClient#30.4.21.1@SHB-L012071.
[2020-03-12 11:58:13]Role initialization failure for ServiceName: LdapClient.
[2020-03-12 11:58:13]Service initialization failure for CLUSTER[name: huawei].

日志报错信息： /bin/systemctl restart  sssd.service 重启失败

[2020-03-12 00:45:48] ERROR: [ldapclient_install.sh:665]: SlapdClient  SHB-L012071  restart sssd failed!
Redirecting to /bin/systemctl restart  crond.service
Redirecting to /bin/systemctl restart  crond.service
2020-03-12 00:45:49 749 SHB-L0120794 SlapdClient INFO cron restart success! 
2020-03-12 00:45:49 790 SHB-L0120794 SlapdClient INFO sssd monitor  already stopped! 
Redirecting to /bin/systemctl stop  nscd.service
Redirecting to /bin/systemctl stop  sssd.service
SSSD is already running
[2020-03-12 00:45:50] ERROR: [ldapclient_install.sh:971]: SlapdClient  SHB-L012071  Install rpm or generate config files failed.

解决方案：先停止LADP的服务，再删除sssd.pid文件

# 1. 添加文件
cd /etc/sssd
touch sssd.conf


# 2. 停止LDAP服务
#停服务
systemctl stop slapd 
#服务生效
systemctl enable slapd
#查看状态
systemctl status slapd


# 3. 删除sssd.pid
rm –rf /var/run/sssd.pid -y





# 4. 停止LDAP服务日志信息
[root@SHB-L0120796 sssd]# systemctl stop slapd
[root@SHB-L0120796 sssd]# systemctl enable slapd
[root@SHB-L0120796 sssd]# systemctl status slapd
● slapd.service - OpenLDAP Server Daemon
   Loaded: loaded (/usr/lib/systemd/system/slapd.service; enabled; vendor preset: disabled)
   Active: inactive (dead) since Thu 2020-03-12 12:05:32 CST; 10s ago
     Docs: man:slapd
           man:slapd-config
           man:slapd-hdb
           man:slapd-mdb
           file:///usr/share/doc/openldap-servers/guide.html
 Main PID: 10829 (code=exited, status=0/SUCCESS)

Mar 12 11:59:47 SHB-L012071 runuser[10823]: pam_unix(runuser:session): session opened for user ldap by (uid=0)
Mar 12 11:59:47 SHB-L012071 runuser[10823]: pam_unix(runuser:session): session closed for user ldap
Mar 12 11:59:47 SHB-L012071 slapd[10828]: @(#) $OpenLDAP: slapd 2.4.44 (Jan 29 2019 17:42:45) $
                                                   [e-mail]:/builddir/build/BUILD/o...slapd
Mar 12 11:59:48 SHB-L012071 slapd[10829]: slapd starting
Mar 12 11:59:48 SHB-L012071 systemd[1]: Started OpenLDAP Server Daemon.
Mar 12 12:05:32 SHB-L012071 systemd[1]: Stopping OpenLDAP Server Daemon...
Mar 12 12:05:32 SHB-L012071 slapd[10829]: daemon: shutdown requested and initiated.
Mar 12 12:05:32 SHB-L012071 slapd[10829]: slapd shutdown: waiting for 0 operations/tasks to finish
Mar 12 12:05:32 SHB-L012071 slapd[10829]: slapd stopped.
Mar 12 12:05:32 SHB-L012071 systemd[1]: Stopped OpenLDAP Server Daemon.

 

问题3. LdapClient 安装失败 

报错信息：Failed to install HA. (main)

[2020-03-12 12:44:41] INFO [postinstall.sh:457] Start install HA. (installHA)
[2020-03-12 12:44:41] INFO [postinstall.sh:458] g_dbSoftPath:/opt/huawei/Bigdata/FusionInsight_BASE_V100R002C70SPC200/install/FusionInsight-dbservice-2.7.0/package,HA_PACKAGE_PATTERN:HA-*.tar.gz (installHA)
[2020-03-12 12:44:41] INFO [postinstall.sh:465] HA packageName is : HA-1.1.8.9-RHEL6.4-x86-64. (installHA)
[2020-03-12 12:44:41] ERROR [postinstall.sh:477] Failed to install ha. cmd is:/opt/huawei/Bigdata/FusionInsight_BASE_V100R002C70SPC200/install/FusionInsight-dbservice-2.7.0/package/HA-1.1.8.9-RHEL6.4-x86-64/install.sh -p /opt/huawei/Bigdata/FusionInsight_BASE_V100R002C70SPC200/install/FusionInsight-dbservice-2.7.0 -u omm -g wheel (installHA)
[2020-03-12 12:44:41] ERROR [postinstall.sh:670] Failed to install HA. (main)

解决方案：更新FusionInsight_Manager文件夹中的HA文件,默认路径为：/opt/huawei/Bigdata

#执行以下命令解压文件 
cd /opt/FusionInsight_Manager/software/om/package/

#解压文件
tar -xzf om.tar.gz

tar -xf om/omm/OMS-V100R001C00.x86_64.tar -C om/omm

#b.解压HA压缩文件
tar -xzf om/omm/OMS-V100R001C00.x86_64/package/software/HA-1.1.8.9-RHEL6.4-x86-64.tar.gz -C om/omm/OMS-V100R001C00.x86_64/package/software/

#使用更新文件，替换HA压缩文件解压后的文件  
cp -arp /opt/huawei/Bigdata/ha.bin om/omm/OMS-V100R001C00.x86_64/package/software/HA-1.1.8.9-RHEL6.4-x86-64/software/module/hacom/bin

#/opt/huawei/Bigdata/ha.bin文件不存在就创建
touch /opt/huawei/Bigdata/ha.bin


#执行
cd om/omm/OMS-V100R001C00.x86_64/package/software/

#将替换后的HA文件全部重新压缩。 
tar -czf HA-1.1.8.9-RHEL6.4-x86-64.tar.gz HA-1.1.8.9-RHEL6.4-x86-64

#删除临时文件
rm -rf HA-1.1.8.9-RHEL6.4-x86-64

问题4：ERROR:The database is not running

报错信息：

[2020-03-11 17:28:29] INFO Succeed to install oms. [install.sh(install_oms):1242](12124)
[2020-03-11 17:28:29] DEBUG Check whether db is running. [install.sh(is_db_running):2054](12124)
[2020-03-11 17:28:29] ERROR The database is not running. [install.sh(is_db_running):2058](12124)
ERROR:The database is not running.
[2020-03-11 17:28:29] ERROR Installation failed. For details about the error, see the log file /var/log/Bigdata/controller/scriptlog/install.log. [install.sh(post_install):500](708)

解决方案：（无解）

重装centos系统

问题5：首次登陆，无法修改密码的错误

解决方案：

启动ldap服务

yum -y install openldap openldap-servers openldap-clients

cp /usr/share/openldap-servers/DB_CONFIG.example /var/lib/ldap/DB_CONFIG
chown ldap:ldap /var/lib/ldap/DB_CONFIG

#首次安装修改密码错误的情况下启动ldap服务
systemctl start slapd 
systemctl enable slapd
systemctl status slapd

问题6：操作失败。

重试创建集群失败。详细错误信息如下：Failed to complete setting up of newly allocated 2 node(s) with agent/JDK.

报错信息：

[2020-03-11 20:51:10]Begin to install nodeagent.
[2020-03-11 20:51:58]Operation success for Nodeagent#30.4.21.17@SHB-L012071.
[2020-03-11 20:52:47]Operation failure for Nodeagent#30.4.21.14@SHB-L012072.
[2020-03-11 20:52:58]Operation failure for Nodeagent#30.4.21.11@SHB-L012073.
[2020-03-11 20:52:58]Begin to update hosts file.
[2020-03-11 20:53:04]Update hosts file complete.
[2020-03-11 20:53:04]Failed to install nodeagent.

解决方案：卸载服务

#执行以下脚本卸载agent。

$BIGDATA_HOME/om-agent/nodeagent/setup/uninstall.sh

#omm密码：Bigdata123@

问题7：dbservice 的HA安装失败

报错信息：

[2020-03-12 13:16:04] INFO [postinstall.sh:465] HA packageName is : HA-1.1.8.9-RHEL6.4-x86-64. (installHA)
[2020-03-12 13:16:05] ERROR [postinstall.sh:477] Failed to install ha. cmd is:/opt/huawei/Bigdata/FusionInsight_BASE_V100R002C70SPC200/install/FusionInsight-dbservice-2.7.0/package/HA-1.1.8.9-RHEL6.4-x86-64/install.sh -p /opt/huawei/Bigdata/FusionInsight_BASE_V100R002C70SPC200/install/FusionInsight-dbservice-2.7.0 -u omm -g wheel (installHA)
[2020-03-12 13:16:05] ERROR [postinstall.sh:670] Failed to install HA. (main)

解决方案：更新dbservice 的HA

# 执行以下命令切换到目录
cd /opt/FusionInsight_Manager/software/packs/

# 解压安装文件。
tar -xzf FusionInsight_BASE/software/components/package/FusionInsight-dbservice-2.7.0.tar.gz -C FusionInsight_BASE/software/components/package/

# 解压HA压缩文件
tar -zxf FusionInsight_BASE/software/components/package/package/HA-1.1.8.9-RHEL6.4-x86-64.tar.gz -C FusionInsight_BASE/software/components/package/package

# 替换HA压缩文件解压后的文件
cp -arp /opt/huawei/Bigdata/ha.bin FusionInsight_BASE/software/components/package/package/HA-1.1.8.9-RHEL6.4-x86-64/software/module/hacom/bin/

# 执行
cd FusionInsight_BASE/software/components/package/package

# 替换后的HA文件全部重新压缩。
rm -rf HA-1.1.8.9-RHEL6.4-x86-64

# 执行 
  cd ../

# 然后将替换后的DBService文件全部重新压缩，并删除临时文件
tar -zcf FusionInsight-dbservice-2.7.0.tar.gz adapter package sbin security setup share tools

# 删除解压后的文件
rm -rf adapter package sbin security setup share tools

# 执行
  cd /opt/FusionInsight_Manager/software/packs/

# 将替换后的基础包文件全部重新压缩，并删除临时文件。
  tar -zcf FusionInsight_BASE_V100R002C70SPC200_RHEL.tar.gz FusionInsight_BASE

# 删除解压后的文件
  rm -rf FusionInsight_BASE