[database] gbase 8A MPP cluster v95 installation and uninstall

GBase 8a MPP Cluster V95 load and unload

Catalog

One 、 Cluster node environment design

Two 、 Preparations before installation

3、 ... and 、 Install cluster

1、 Create on all nodes of the cluster DBA user

2、 Create the installation directory on all nodes of the cluster and authorize

3、 Installation package decompression （/opt Under the table of contents ）

4、 Set the environment variable

5、 Modify the installation configuration file of the master node （demo.options）

6、 Execute installation script

7、 Check the cluster status

8、 Apply for authorization

9、 Import and check authorization

10、 Start all cluster services on all nodes of the cluster

11、 Set slice information （ Create publication ）

12、 Database initialization

13、 Create library table

Four 、 Uninstallation of cluster

5、 ... and 、 common problem

1. Execute... On the management node gcadmin, System prompt

2.SSH Service confirmation 22 The port is disabled , modify 8a Cluster port

---

One 、 Cluster node environment design

IP	role	OS
10.10.14.101	management 、 data 、 Master node	RedHat7.3
10.10.14.102	management 、 Data nodes	RedHat7.3
10.10.14.103	management 、 Data nodes	RedHat7.3

Network requirements ： Each node IP It's the same segment , And can connect with each other ; Turn on SSH service ; Turn off firewall 、 close seLinux service .

Two 、 Preparations before installation

Turn off firewall or Open the corresponding port

systemctl disable firewalld

systemctl status firewalld

● firewalld.service - firewalld - dynamic firewall daemon Loaded: loaded (/usr/lib/systemd/system/firewalld.service; disabled; vendor preset: enabled) Active: inactive (dead) Docs: man:firewalld(1）

close selinux service

[[email protected] opt]# cd /etc/sysconfig
[[email protected] sysconfig]# vi selinux

# This file controls the state of SELinux on the system.
# SELINUX= can take one of these three values:
# enforcing - SELinux security policy is enforced.
# permissive - SELinux prints warnings instead of enforcing.
# disabled - No SELinux policy is loaded.
#SELINUX=enforcing
SELINUX=disabled
# SELINUXTYPE= can take one of three two values:
# targeted - Targeted processes are protected,
# minimum - Modification of targeted policy. Only selected processes are protected.
# mls - Multi Level Security protection.
SELINUXTYPE=targeted

After restarting the node ：

[[email protected] opt]# sestatus

SELinux status: disabled

3、 ... and 、 Install cluster

1、 In clusters all Create... On node DBA user

# useradd gbase
# passwd gbase

2、 In clusters all Create the installation directory on the node and authorize

# mkdir -p /opt/gbase
# chown gbase:gbase /opt/gbase
# chown gbase:gbase /tmp

3、 Installation package decompression （/opt Under the table of contents ）

[[email protected] opt]$ tar xfj GBase8a_MPP_Cluster-License-9.5.2.39-redhat7.3-x86_64.tar.bz2

After decompression is complete ,opt You can see the generation gcinstall The installation directory .

4、 Set the environment variable

    Copy the environment setting script of the master node （SetSysEnv.py） to 2 platform From the node

  # scp [email protected]:/opt/gcinstall/SetSysEnv.py /opt
  # scp [email protected]:/opt/gcinstall/SetSysEnv.py /opt

    function 3 Of nodes SetSysEnv.py Script configuration installation environment

  # python SetSysEnv.py --dbaUser=gbase --installPrefix=/opt/gbase --cgroup
  # python SetSysEnv.py --dbaUser=gbase --installPrefix=/opt/gbase --cgroup
  # python SetSysEnv.py --dbaUser=gbase --installPrefix=/opt/gbase --cgroup

Error occurs when setting environment variables

Solution

The lack of bc, install bc

yum install bc

 y y

5、 Modify the installation configuration file of the master node （demo.options）

    Switch to gbase user ：
    [[email protected] gcinstall]$ su - gbase
    [[email protected] gcinstall]$ vi demo.options

installPrefix = /opt/gbase
coordinateHost =  10.10.14.101,10.10.14.102,10.10.14.103
coordinateHostNodeID = 101,102,103
dataHost = 10.10.14.101,10.10.14.102,10.10.14.103
#existCoordinateHost =
#existDataHost =
dbaUser = gbase
dbaGroup = gbase
dbaPwd = '#310619edc'
rootPwd = 'root'
#rootPwdFile = rootPwd.json

dbaPwd yes gbase Password of the account
rootPwd yes root Password of the account
 

6、 Execute installation script

Before executing the installation script 3 Node installation bzip2

yum -y install bzip2

Execute the installation script on the master node

./gcinstall.py --silent=demo.options

remarks ： During installation , First carry out environmental inspection , There may be a mistake , List the missing rpm Dependent package name , It means that the operating system is not completely installed rpm package , Need basis rpm The name of the package goes to each node to install one by one .
8a List of required mandatory dependency packages , Please check the installation directory gcinstall Under the dependRpms file

$ cat /opt/gcinstall/dependRpms

pcre
krb5-libs
libdb
glibc
keyutils-libs
libidn
libuuid
ncurses-libs
libgpg-error
libgomp
libstdc++
libcom_err
libgcc
python-libs
libselinux
libgcrypt
nss-softokn-freebl

During the execution of the installation script , If the system prompts that rpm Dependency package , You need to follow the above rpm List name , Install the missing packages on each node .

Install all required dependencies

yum install xx

Be careful ： If the operating system does not cgroup（ Resource management ） Components , The following warning will appear , Input Y You can enter. .

y y

7、 Check the cluster status

After execution , Switch to gbase user , View the cluster status

gcadmin

Because there is no registration authorization ,gcluster and gnode Service is close The state is normal .

Be careful ： If the system prompts that the command cannot be found , Switch operating system accounts     $ su gbase

If there is permission error , Need to match ssh Password free login

8、 Apply for authorization

    Be careful ： If you use the cloud server provided ,/opt/ Authorization file under directory （*.lic） You can use it directly . Please skip. “ Apply for authorization ” step .

① Export the fingerprint information of each node of the cluster ：

[173]$ ./gethostsid -n 10.10.14.101,10.10.14.102,10.10.14.103 -u root -p password -f /tmp/finger.txt

② Apply for authorization

    Email to ：[email protected]; Copy to [email protected]; The attachment is the fingerprint information file finger.txt.
    Email title ：GBase 8a MPP Cluster v95 license apply
    Message body ：

1. Customer name : Your full name
2. Project name : 2022 year X month GBase 8a MPP Cluster GDCA Certification training
3. Applicant : Your name
4. The reason for applying : Training practice
5. The period of validity : 3 Months
6. Operating system name and version : CentOS Linux release 7.6.1810 (Core)
7. 8a Cluster version : GBase8a_MPP_Cluster-License-9.5.2.39-redhat7.3-x86_64.tar.bz2

③ The processing time point of authorization application is working day 9:30、13:30 and 17:30. Delegates receive authorization documents (*.lic) After uploading to the master node /tmp Next .

9、 Import and check authorization

① Import authorization ：

[[email protected] gcinstall]$ ./License -n 10.10.14.101,10.10.14.102,10.10.14.103 -f /tmp/20220722-19.lic -u gbase -p '#310619edc'

② Check the import of authorization ：

[[email protected] gcinstall]$ ./chkLicense -n 10.10.14.101,10.10.14.102,10.10.14.103 -u gbase -p '#310619edc'

License Status description ：

    is_exist Used to identify license Does the file exist ： yes On behalf of there ,no nonexistence ;
    version Used to identify license type ： trial For trial version , business Commercial version ;
    expire_time Used to identify the trial version license Due date for , Only testing the trial version license Only when ;
    is_valid Used to identify license Whether it works ： yes representative license It works ,no representative license invalid ;

    If the authorization is found to be invalid （is_valid is no）, It may be due to the hardware change of the cluster node , Please regenerate the fingerprint file and apply for authorization by email .

10、 In clusters all Start all cluster services on the node

$ su - gbase
$ gcluster_services all start

Be careful ： All nodes are started
View the cluster status ：

$ gcadmin

CLUSTER STATE: ACTIVE
VIRTUAL CLUSTER MODE: NORMAL

11、 Set slice information （ Create publication ）

[[email protected] gcinstall]$ gcadmin distribution gcChangeInfo.xml p 2 d 1 pattern 1

• gcinstall Lower generation gcChangeInfo.xml file

  Check the cluster status again    gcadmin

  You can also execute the following commands to view the release information

$ gcadmin showdistribution node

12、 Database initialization

    Execute the following command on the management node （ database root The password is empty by default ）

    [173]$ gccli -u root -p
    gbase> initnodedatamap;

13、 Create library table

Successful initialization , Then the whole 8a Cluster installation complete , You can create the first library and table .

gbase> create database test;
gbase>show databases;
gbase>use test;
gbase>create table t(id int ,name varchar(20));
gbase>show tables;

Four 、 Uninstallation of cluster

stop it all All cluster services of the node

[[email protected] tmp]$ gcluster_services all stop

Execute the uninstall command on the master node

[[email protected] gcinstall]$ ./unInstall.py --silent=demo.options

5、 ... and 、 common problem

1. Execute... On the management node gcadmin, System prompt

Could not initialize CRM instance error: [122]->[can not connect to any server]

    reason ： The cluster service of all nodes is not started
    resolvent ：
    [173]$ gcluster_services all start
    [174]$ gcluster_services all start
    [175]$ gcluster_services all start

2.SSH Service confirmation 22 The port is disabled , modify 8a Cluster port

    resolvent ：

    modify SSH The configuration file .
    [173]# cd /etc/ssh
    [173]# vi ssh_config
    hypothesis , Modify profile “Port” The value of is 10022
    restart SSH service ：
    [173]#service sshd restart
    see SSH Whether the listening port is modified to 10022
    [173]# netstat -tunlp | grep ssh
    Turn off all management node cluster services
    [173]$ gcluster_services all stop
    [174]$ gcluster_services all stop
    [175]$ gcluster_services all stop
    Modify all management nodes $GCWARE_BASE/config/gcware.conf Medium
    gcware Profile's node_ssh_port: 22
    Restart all management node cluster services
    [173]$ gcluster_services all start
    [174]$ gcluster_services all start
    [175]$ gcluster_services all start

---

End of the flower ！！！

        Friends, this is the end of our article , The content of this article is not difficult , If there are any friends who can't, we can talk together ~

         If your friends feel that the content of the article is helpful to you , Sure give the thumbs-up , Follow articles and columns as well as Pay attention to me , Hey, hey, I'll update the article regularly , Thank you for your support ！