At least 1 server is required to run all the software required for e-mail, messaging and code services.
However, if this one server breaks then at least services would be interrupted until a new server is
started up, all the software is installed and data recovered.
In the worse case scenario, some data would be lost since the last backup.
Without redundancy, recovery time, after the server breaks can take a few hours to 1 day and requires data recovery from backups.
To improve reliability and prevent service interruption we have redundancy, that is an additional server running all the software and duplicating all the data. In a system with redundancy active, even if 1 server breaks, the other still works and can automatically take over all the services and users' requests.
In systems with redundancy active there is no data loss if one of the server breaks. Typically, there is
no service interruption, with users simply having to reconnect to continue working. Sometimes, in the
worse case, the interruption can last a few seconds.
With redundancy, recovery time, after server breaks takes from 0 to a few seconds and is fully automatic with no need to recover data from backups.