Virtualisation Tips & Tricks

This is fastest method using Postgres repmgr to clone DB from primary cell. Detailed explanation is here https://blog.raveland.tech/post/postgresql_repmgr_pgbouncer_en/

Stop Postgres DB:

systemctl stop vpostgres

Remove the content of the pgdata on the standby nodes:

rm -rf /var/vmware/vpostgres/current/pgdata

Copy the pgdata content form the primary node to the standby nodes:

su - postgres
/opt/vmware/vpostgres/current/bin/repmgr -h <primary_database_IP> -U repmgr -d repmgr -f /opt/vmware/vpostgres/current/etc/repmgr.conf standby clone

Make sure that the vpostgres service is running on the standby cell  if down, restart with command:

systemctl start vpostgres
systemctl status vpostgres

Re-register the standby nodes:

/opt/vmware/vpostgres/current/bin/repmgr -h <primary_database_IP> -U repmgr -d repmgr -f /opt/vmware/vpostgres/current/etc/repmgr.conf standby register --force

After recovery failover mode will be in failed state. It is unfortunately not possible to change the failover method to automatic from the Cloud Director web console. The only way to change this is by using the API from Cloud Director.

• Configure the root credentials of the Cloud Director cell in the authorization tab.
• Accept: application/json

https://10.10.10.1:5480/api/1.0.0/nodes/failover/automatic

Health Check

Got script collection from VMware technical support which will be included in next vCloud Director release.

Download attached .tar file and then upload to server. Execute commands described below:

tar -zxvf serviceability2022-07-04_15_UTC.tar.gz
root/service/setup.sh
source ~/.bash_profile
 
Then you can check cluster health from root commandline:
repmgr_cluster_show
repmgr_cluster_check
repmgr_cluster_matrix
repmgr_cluster_mevent
repmgr_cluster_event
env_check

Check Replication and Failover in a Cluster

systemctl status repmgrd.service
 
systemctl show repmgrd.service | grep -e "Environment\|ExecStart"