May I check following questions with you:
We got two separate physical servers set up, Server1 and Server2 in one site separately connected with high speed LAN. Both of them are Windows 2012 + SQL Server 2014 + hMailServer. The purpose is to set Server1 active, and Server2 standy.
Imaging there is an running email broadcasting covering about 1000 emails (not junk emails), and somehow the Server1 corrupts, and we have to switch it to Server2 and continue the on-going email broadcasting.
Question1.1 for this scenario is, besides following preparation:
- 1.Copy hMailServer Configuration.
2.The database and hMailServer Data Directory have been synchronized from Server1 to Server2.
3.Check that the Windows Firewall is not blocking traffic to hMailServer.
4.Update any MX record which was pointing at the old host name.
5.Update the host name in the hMailServer configuration (if it has changed).
Question1.2 for this scenario is:
How to ensure the Server2 can continue automatically sending the remaining emails(which are not delivered yet from Server1), instead of re-sending all the 1000 emails from the beginning. If the SQL Server Replication(to backup the database) , and the built-in backup of hMailServer(to backup the Data Directory) is enough to secure no latency which may cause the inconsistency of the email queue?
Or it is not possible technically?
Same configuration as Scenario 1, but both the Server1 and Server2 will function actively sending broadcast emails(not junk emails again). However, to ease the load pressure of each server, Server1 will serve for Email Group 1, and Server2 will serve for Email Group 2. In the case Server1 fails, the Server2 needs to take over the remaining task of Email Group 1(namely the Server2 will need to not only proceed the Email Group 2 but also continue the Email Group 1 for those undelivered emails).
Question2.1 for this scenario is:
Technically is that possile? If possible, what needs to configured, or what script needs to be prepared (which can be run when the cluster identify the Server1 is down).
Appreciate your opinions in advance.