We are planning to upgrade to sql 2016 enterprise edition and at the same time set up fail over cluster. Here's the basic requirements.
A load server where all the daily etls run to load data from OLTP to OLAP (data warehouse).
A reporting server - after the daily load on the load serve a differential backup is taken and restored to the reporting server. The end users and report portal connects to this server.
Please advise what makes sense for fail over clustering. I read a lot of stuff on always on high availability and fail over clustering, which is only confusing me, so I wanted to put my questions on this forum to get some simple answers.
I assume we can do an active passive fail over clustering for the reporting server, which if goes down, the passive node take over and the reports continue to show valid most current data seamlessly. I assume the passive node is read only. in this architecture, I understand we do not require 2 licenses of sql server for the reporting server fail over cluster. Please correct me if I am wrong.
Do we also have an option to have a similar configuration for the load server, where if during the load, the active node goes down, the passive node takes over and the etl continues seamlessly loading on the passive node, until the active node comes back up again. Is this possible or just my imagination? if possible, is this active active configuration, do we need two copies of SQL server license?
Do these configurations need shared storage or individual storage?
Any advise is highly appreciated.
Thanks in advance.