Hot Replacement of TSS servers

Sometimes, the servers on which TSS are installed must be replaced. Because an issue had not been handled on time and grown out of control or simply because hardware or software must be upgraded.

The following procedure is meant to avoid any service interruption during the operation, but be careful that a few minutes of data may be lost during the operation. This includes call logs as well as every configuration changes that may occurs during this period.

1.Set the first server offline

If one of the servers is not working properly or has an out-of-sync database, it will be the first server to be set offline. If not, the first server to be set offline will be the primary.

Connect to this server and shut IIS service down, so all devices switch to the other one.

2.Wait for all database operations to be replicated on the other server

On SQL Server, use the replication monitor (see Replication)

On MySQL, monitor it using show slave status on the second server (Slave_SQL_Running_State must state it has read all relay log)

3.Cleanup or replace the server

Stop the replication and either wipe out everything that is problematic, or set the server offline and replace it with a new one.

•The new server must be accessible using the same uri as the previous one (ip address and/or FQDN)

•The same version(s) of TSS firmware(s) must be installed on both first and second servers

•IIS service must not be started until server is ready and database is replicated

4.Take a database backup on the second server

5.Restore backup on the first server

6.Synchronize databases (if possible) and switch servers

1.Set up replication between both first and second servers (if possible)

2.Start IIS service on first server

3.Stop IIS service on the second server

7.Wait for all database operations to be replicated on the other server

On SQL Server, use the replication monitor (see Replication)

On MySQL, monitor it using show slave status on the first server (Slave_SQL_Running_State must state it has read all relay log)

8.Cleanup or replace the server

Stop the replication and either wipe out everything that is problematic, or set the server offline and replace it with a new one.

•The new server must have be accessible using the same uri as the previous one (ip address and/or FQDN)

•The same version(s) of TSS firmware(s) must be installed on both first and second servers

•IIS service must not be started until server is ready and database is replicated

9.Take a database backup on the first server

10.Restore backup on the second server

11.Re-set replication

12.Start IIS service on the second server