An automation performs the following on a monthly basis: - Disable accounts that have not logged on in the last 60 days - Remove firewall rules with zero hits in the last 30 days - Remove DNS entries

Disabling accounts might affect access, but it's less likely to directly cause a replication failover test to fail unless service accounts specifically for replication were inactive and disabled.

Removing firewall rules with zero hits in the last 30 days is the most likely cause because quarterly replication failover testing means the associated network traffic might only occur once every three months. If the necessary firewall rules did not register any hits within the 30-day window before automation, they would be removed, thus preventing the failover test from succeeding.

Removing DNS entries could prevent hostname resolution, but replication often uses direct IP communication or specific service configurations not solely dependent on dynamic DNS updates within a 60-day window for failover paths.

Archiving logs does not directly impact the operational functionality or network connectivity required for replication failover; it only moves historical data for storage.