Exchange Server DAG DR Site – Failover Duration and Recommendations

Question

Hi everyone,

We are implementing a Disaster Recovery (DR) strategy for our on-premises Exchange Server 2019 environment and would like to clarify some points regarding failover and DR site operation.

Environment:

Exchange DAG with 6 Mailbox servers:

3 in HQ (Primary site)

  3 in DR site
  
  Planning to use a third site for the File Share Witness.
  
  DAG replication and health are monitored and maintained.
  
  We are preparing for both **planned failovers (e.g., DR drills)** and **unplanned failovers (e.g., HQ outage)**.
  

Key Questions:

How long does it typically take for the DR site to become the active site in case of:

A planned failover (DR drill)?

  An **unplanned failover (HQ outage)?**
  
  What are the recommended **maximum durations** for running the DR site as the active site before failing back to HQ?
  
  Are there any **technical risks or limitations** when operating from the DR site for an extended period (days/weeks)?
  
  What are the **recommended RTO and RPO** values for such a setup?
  

1. Any best practices or things to consider regarding:

DAG quorum and File Share Witness across three sites

Failback procedures

• Namespace or DNS adjustments during failover/failback

Answer 1

If you have a 3rd site for the FSW, then you dont have a DR scenario. You have two peer data centers and HA. I would design and build with that in mind. In other words, treat both data centers the same and allow databases to be mounted on any server in either data center.

If you build with that in mind, then you wont have to worry about data center failures and its easy to test and prove out.