NodeChef supports replication from the Master to no more than two standby nodes. If you require more standby nodes, please open a ticket from the dashbooard.
NodeChef configures synchronous replication with "remote_write" commit to one of your standby servers. When the master node fails, this node becomes the next master node.
You can configure replication when deploying the server or after deploying the server. To configure replication when deploying the database, select the number of standby servers from the "Standby servers" field. You can always change the number of standby servers at any time.
To change the number of standby servers, navigate to Deployments → Deployed Apps from the dashboard. Find your database listed under the table. Under the database column, select the number of nodes that will make up your cluster. The number of nodes includes the master node. Thus if you select 2, you will end up with a master and a standby but selecting 3, you will end up with one master and 2 standby nodes and so on.
After selecting the number of nodes, click on Cluster Actions and click the Reconfigure option to initiate the process. Please note your account will be billed immediately after the resources are provisioned.
You can always monitor the state of replication to standby servers from the task manager. Click on PGDB Actions → Replication.
The table displayed in the replication panel contains the columns SendLag, WriteLag and FlushLag.
The difference between the masters current transaction log position and the last transaction log position sent to this standby server in bytes, kilobytes or megabytes.
The difference between the masters current transaction log position and the last transaction log position written to disk by this standby server in bytes, kilobytes or megabytes.
The difference between the masters current transaction log position and the last transaction log position flushed to disk by this standby server in bytes, kilobytes or megabytes.