HA Upgrade from Version 11.4

If the current system is using the HA feature and it is version 11.4, then the following procedure must be followed, instead of using Applying Patches & Upgrades With HA.

Prognosis Server upgrade requires the PostgreSQL Schema to be independently upgraded to the latest schema, for this to occur correctly, replication needs to be stopped between the HA Pair and promote the standby database to a master database before upgrading.

Follow the steps below to upgrade from version 11.4 to 11.9.

There may be data loss while the upgrade procedure is being conducted due to the loss of replication

Disable HA

Login to both HA Pair servers using the Web Application - Administration tool or the Windows Client, and edit the HA Configuration to disable the High Availability feature

ENABLE HA (N)

Restart the HA Configuration on both servers

Promote the Secondary server

On the Secondary server, start a Command Prompt, locate the following folder and execute the command:

> cd <Prognosis_Home>\Server\Configuration\utils\ha
> master.bat <master_ip> <standby_ip>

This will promote the Secondary server to master and stop replication.

Verify that PostgreSQL Replication has stopped

Execute the command to verify the Primary server has stopped replication:

> psql -U prognosis -h <master_ip>
PSQL> SELECT pg_is_in_recovery();
PSQL> exit

This should return f (False).

Then execute the commands to verify the Secondary server has stopped replication:

> psql -U prognosis
PSQL> SELECT pg_is_in_recovery();
PSQL> exit

This should return f (False).

When both PostgreSQL server show false for in recovery, then replication has been stopped.

Patch/Upgrade both HA Pair servers

Follow the standard operating procedures to either patch or upgrade Prognosis Server on both Primary and Secondary servers.

Promote the Primary server to master

Start a command prompt on the Primary server and execute:

> cd <Prognosis_Home>\Server\Configuration\utils\ha
> master.bat <master_ip> <standby_ip>

Setup the Secondary server as standby

Start a command prompt on the Secondary server and execute:

> cd <Prognosis_Home>\Server\Configuration\utils\ha
> standby.bat <master_ip> <standby_ip>

Disable HA

Login to both HA Pair servers using the Web Application - Administration tool or the Windows Client, and edit the HA Configuration to enable High Availability feature

ENABLE HA (Y)

Restart the HA Configuration on both servers

Verify that PostgreSQL Replication has started

Repeat Step 3, but this time the pg_is_in_recovery() result should be true (t) on the Secondary server only.






























Provide feedback on this article