Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 4 Current »

Steps for creating Data pool:

Literature

  1. Master Server/ Master: MySQL database running on production servers.

  2. Slave Server/ Slave: MySQL database running in the Datapool.

  3. Docker image: Image pulled from Docker Hub, that is used to spin up docker containers.

  4. Docker container: Instance of a docker image.

  5. Data pool: Slaves(mysql instances) are running here.

Implementation Stack

  1. MySQL Master Slave Replication

  2. Docker

  3. Docker image: mysql/mysql-server:5.7

  4. Systemd Service

  5. Systemd Timer


Perform the following steps on Master Server that is to be replicated on the Data pool:

  1. Change to root user.

    1. sudo su

  2. Change directory to /etc/mysql/mysql.conf.d

    1. cd /etc/mysql/conf.d

  3. Edit mysqld.cnf

mysqld.cnf

press i

comment the following line

bind-address = 127.0.0.1

uncomment the following line

server-id = 1

add the following lines

log_bin = /var/log/mysql/mysql-bin.log

binlog_format = row

binlog_do_db = openmrs

press esc

press :

enter wq and enter

4. Restart mysql service

systemctl restart mysql

5. Login to mysql using root

mysql -u root -p

6. Create a new user

create user ‘<username>’@’<ip_address_of_datapool>’ identified by ‘<password>’;

7. Grant replication slave rights to this user.

grant replication slave on *.* to ‘<username>’@’<ip_address_of_datapool>’;

8. Flush privileges

flush privileges;

9. Flush tables with read lock

flush tables with read lock;

10. Note down the master details - file name and log position(log pos)

show master status;

11. Open a duplicate session and in this new session run the following command

mysqldump -u root -p openmrs > 1openmrs.sql

12. Switch back to the mysql session and run the following commands

unlock tables;

quit;

13. Transfer the file 1openmrs.sql file to the datapool server

Installing Docker Community Edition on the Datapool Server

  1. Add the GPG key for the official Docker Repository system

curl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -

2. Add the docker repository to the apt sources

sudo add-apt-repository "deb [arch=amd64] \ https://download.docker.com/linux/ubuntu $(lsb_release -cs) stable"

3. Update the package database with Docker packages

sudo apt-get update

4. Install Docker

sudo apt-get install -y docker-ce

5. Check whether Docker is running

systemctl status docker

6. Add user to the docker user group to avoid typing sudo

sudo usermod -aG docker ubuntu

7. Check if user is added to the group

id -nG



Install MySQL image using Docker

  1. Download MySQL Server Docker image

docker pull mysql/mysql-server:5.7

2. Check whether the image has been added

docker images


Create MySQL Slave servers on Datapool to replicate the Master servers

  1. Create the following directories

datadir (should be empty)

etc

init

mkdir datadir

mkdir etc

mkdir init

2. Change directory to etc

cd etc

3. Create the file my.cnf

touch my.cnf

4. Edit the my.cnf file

vi my.cnf

press i

Add the following lines in the file

[mysqld]

user=mysql

server-id=2

log-bin=mysql-bin

binlog-format=row

relay-log=mysql-relay

replicate-rewrite-db=openmrs-><name_of_the_database_in_the_slave_mysql>

replicate-do-db=<name_of_the_database_in_the_slave_mysql>

binlog-do-db=<name_of_the_database_in_the_slave_mysql>

press esc, enter :wq, enter

5. Change directory to init

cd ../init

6. Copy the 1openmrs.sql file

7. Edit the 1openmrs.sql file as follows

Add the following line as the first line in the file

USE <name_of_the_database_in_the_slave_mysql>;

Save the file

8. Create 2setup_slave.sql file

touch 2setup_slave.sql

Add the following line

CHANGE MASTER TO MASTER_HOST='<master_ip_address>', MASTER_USER='<slave_user_created_on_the_master_mysql_server>', MASTER_PASSWORD='<slave_user_password>', MASTER_LOG_FILE='<master_log_file_name>', MASTER_LOG_POS=<master_log_pos>;

press esc, enter :wq, enter

9. Use docker to create the mysql server container that will have the above configurations with the following command

docker run --name <container_name> \

-e MSQL_ROOT_PASSWORD=<root_password> \

-e MYSQL_USER=<slave_user> \

-e MYSQL_PASSWORD=<slave_user_password> \

-e MYSQL_DATABASE=<database_name_that_will_be_populated_with_master_data> \

--mount type=bind,source=<location_to_datadir>,target=/var/lib/mysql \

--mount type=bind,source=<location_to_my.cnf>,target=/etc/my.cnf \

--mount type=bind,source=<location_to_init>,target=/docker-entrypoint-initdb.d \

-p <host_port_to_be_mapped>:3306 \

-d mysql/mysql-server:5.7

10. Check with the following command whether the container has been successfully created

docker logs <container_name>


Create Systemd Service and Timer to sync the slave server with the master server at a particular time

  1. Create the following files /lib/systemd/system

  • <service_name>.service

  • <service_name>.timer

2. Add the following lines to .service file

[Unit]

Description=Service started for syncing with Master afi

[Service]

ExecStart=<location of the script which will be run to sync slave with master>

[Install]

WantedBy=multi-user.target

3. Add the following lines to .timer file

[Unit]

Description=Timer started for syncing with Master afi

[Timer]

OnCalendar=*-*-* <HH:MM:SS>

[Install]

WantedBy=multi-user.target


Create the Script which will sync the slave with the master, and be run by the service.

#!/bin/bash

docker exec <container_name> mysql --user=root --password=<root_password> -e "

STOP SLAVE IO_THREAD;

START SLAVE SQL_THREAD;"

sleep 1m

docker exec <container_name> mysql --user=root --password=<root_password> -e "

STOP SLAVE SQL_THREAD;

START SLAVE IO_THREAD;"

References


https://dev.mysql.com/doc/mysql-installation-excerpt/5.7/en/docker-mysql-getting-started.html

https://www.certdepot.net/rhel7-use-systemd-timers/

https://linuxconfig.org/how-to-create-systemd-service-unit-in-linux

https://severalnines.com/database-blog/multiple-delayed-replication-slaves-disaster-recovery-low-rto

Current Datapool Status

  1. demo is not functioning (disabled)

  2. afi and vso slave databases are accessible via 3307 and 3308 ports

  3. refer the demo file structure and scripts for creating further containers for slaves, debugging and logging

Credentials

  1. AFI

User: root

Password: i10hi1c_afi

User: dbslave

Password: dbslave_afi

Database: OpenMRS_AFI

Port: 3307

SQL URI: mysql://dbslave:dbslave_afi@13.229.131.242:3307/OpenMRS_AFI

2. VSO

User:root

Password:i10hi1c_vso

User:dbslave

Password: dbslave_vso

Database: OpenMRS_VSO

Port:3308

SQL URI: mysql://dbslave:dbslave_vso@13.229.131.242:3308/OpenMRS_VSO


3. Replication Users on AFI and VSO production servers:

  1. AFI

User: dbslave

Password: dbslave_afi

(dbslave@13.229.131.242)

2. VSO

User:dbslave

Password: dbslave_vso

(dbslave@13.229.131.242)

Encountered Issues:

  1. To re-start Docker

To check if the docker container is running, run the following command:

docker ps

<if no row returns → docker is down>

Path: /home/ubuntu/datalake/afi/scripts

Names of the container - afi, vso

Command to start docker:

docker restart <container name>

Once the docker is restarted, it is necessary to restart the sync service & enable the same (else the replication timer will not be scheduled).

The server location of service & timer is: /lib/systemd/system

sync_master_afi.service

sync_master_afi.timer

sync_master_vso.service

sync_master_vso.timer

Command to list the scheduled timers:

system$ sudo systemctl list-timers sync_master*

Command to check the status of the service & timer

systemctl status sync_master_afi.service

systemctl status sync_master_vso.service

systemctl status sync_master_afi.timer

systemctl status sync_master_vso.timer

To enable the service run following commands:

systemctl enable sync_master_afi.service

systemctl enable sync_master_vso.service

To enable the timer run the following commands:

systemctl enable sync_master_afi.timer

systemctl enable sync_master_vso.timer

2. If the datapool is out of sync with production:

Perform the following steps on Master Server that is to be replicated on the Datapool

1.Using Putty, connect to prod server. Change to root user.

sudo su

2. Restart mysql service

systemctl restart mysql

3. Login to mysql using root

mysql -u root -p

4. Run following commands:

RESET MASTER;
FLUSH TABLES WITH READ LOCK;
SHOW MASTER STATUS;

5. The status appears as below: Note that position number varies!

+------------------+----------+--------------+------------------+
| File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
+------------------+----------+--------------+------------------+
| mysql-bin.000001 | 98 | your_DB | |
+------------------+----------+--------------+------------------+

6. Note down the master details - file name and log position(log pos)

7. Open a duplicate session and in this new session run the following command

mysqldump -u root -p openmrs > 1openmrs.sql

8. Switch back to the mysql session and run the following commands

unlock tables;

quit;

9. Transfer the file 1openmrs.sql file to the datapool server

Perform the following steps on Master Server that is to be replicated on the Datapool (In all the steps mentioned below, it is assumed that we are replicating the AFI docker container. For VSO or any other container make the necessary changes)

  1. Stop the docker that is out of sync

    1. docker stop afi

  2. Remove the docker container

    1. docker rm afi

  3. The file structure maintained at datapool is:

  4. Go to path - /home/ubuntu/datalake

  5. Take the backup of the container folder that you want to replace on your local machine

  6. delete the container folder from datalake folder

  7. Now create a new directory for the conatainer

    1. mkdir afi

  8. Under this new folder create following 5 sub folders

    1. mkdir datadir (it should be empty)

    2. mkdir init

    3. mkdir etc

    4. mkdir scripts

    5. mkdir services

  9. Change directory to etc

    1. cd etc

  10. Create the file my.cnf

    1. touch my.cnf

  11. Edit the my.cnf file

    1. vi my.cnf

  12. To insert text press i

    1. Add the following lines in the file:

      [mysqld]

      user=mysql

      server-id=2

      log-bin=mysql-bin

      binlog-format=row

      relay-log=mysql-relay

      replicate-rewrite-db=openmrs->OpenMRS_AFI

      replicate-do-db=OpenMRS_AFI

      binlog-do-db=OpenMRS_AFI

  13. Press esc, enter :wq, enter to save the file

  14. Change directory to init

    1. cd ../init

  15. Copy the 1openmrs.sql file

  16. Edit the 1openmrs.sql file as follows:

    1. Add the following line as the first line in the file

       

      USE <name_of_the_database_in_the_slave_mysql>;

      i.e. USE OpenMRS_AFI;

  17. Press esc, enter :wq, enter to save the file

  18. Create 2setup_slave.sql file

    1. touch 2setup_slave.sql

  19. Add following lines to the sql file:

    1. CHANGE MASTER TO MASTER_HOST='<master_ip_address>', MASTER_USER='<slave_user_created_on_the_master_mysql_server>', MASTER_PASSWORD='<slave_user_password>', MASTER_LOG_FILE='<master_log_file_name>', MASTER_LOG_POS=<master_log_pos>;

    2. in case of afi the the above lines will be

      1. CHANGE MASTER TO MASTER_HOST='13.127.240.201', MASTER_USER='dbslave', MASTER_PASSWORD='dbslave_afi', MASTER_LOG_FILE='mysql-bin.000001', MASTER_LOG_POS=<as noted in step#5>;

  20. Press esc, enter :wq, enter to save the file

  21. Under subfolder “Scripts”, copy the files from backup taken on local machine (under same subfolder) i.e. files viz. create_container.sh, mysql.sh & dbash.sh. Provide the necessary execution rights to the files.

  22. Under subfolder “Services”, create 2 files viz. service.log & debug.txt using command:

    1. touch service.log

    2. touch debug.txt

  23. Use docker to create the mysql server container that will have the above configurations with the following command:

    1. docker run --name <container_name>
      -e MYSQL_ROOT_PASSWORD=<root_password>
      -e MYSQL_USER=<slave_user>
      -e MYSQL_PASSWORD=<slave_user_password>
      -e MYSQL_DATABASE=<database_name_that_will_be_populated_with_master_data>
      --mount type=bind,source=<location_to_datadir>,target=/var/lib/mysql
      --mount type=bind,source=<location_to_my.cnf>,target=/etc/my.cnf
      --mount type=bind,source=<location_to_init>,target=/docker-entrypoint-initdb.d
      -p <host_port_to_be_mapped>:3306
      -d mysql/mysql-server:5.7

    2. In case of afi, the command will be:

      1. docker run --name afi
        -e MYSQL_ROOT_PASSWORD=i10hi1c_afi
        -e MYSQL_USER=dbslave
        -e MYSQL_PASSWORD=dbslave_afi
        -e MYSQL_DATABASE=OpenMRS_AFI
        --mount type=bind,source=/home/ubuntu/datalake/afi/datadir,target=/var/lib/mysql
        --mount type=bind,source=/home/ubuntu/datalake/afi/etc/my.cnf,target=/etc/my.cnf
        --mount type=bind,source=/home/ubuntu/datalake/afi/init,target=/docker-entrypoint-initdb.d
        -p 3307:3306
        -d mysql/mysql-server:5.7 &>> /home/ubuntu/datalake/afi/services/debug.txt

  24. Check with the following command whether the container has been successfully created

    1. docker logs <container_name> i.e. docker logs afi

  25. Run the scripts in “Scripts” folder viz. dbash.sh & mysql.sh

  26. Enable the sync master service & timer as described in Issue#1

  27. Copy the shell scripts from backup (local machine) “Services” folder to “Services” folder in docker container folder i.e /home/ubuntu/datalake/afi

  28. Provide necessary read/write/execute rights to the shell scripts

  29. Login to Superset with admin login

  30. Go to Sources → Databases.

  31. Select the container newly created. Click on Edit record button.

  32. Hit the “Test Connection” button to ensure that the DB connection is re-established.

  33. Check if dashboard displays correct data

 

  • No labels