Data Protection

You can find out long before then: open any copy as a new server and check your application starts, without touching production. The interface never shows a job as successful before it has seen it finish, and it spells out why a failed job failed.

What you lose is the gap between the last successful replication and the moment of failure, measured as RPO. Each replication's RPO is tracked separately, and an alert is raised when the target is missed. The target your servers need is agreed with Nesnet when you choose what to protect; this page doesn't promise a figure.

Deleting a copy has to be confirmed with a short-lived code sent to your verified email address, and even a confirmed delete is held for a while before it takes effect. Copies you have locked cannot be deleted by anyone on the tenant side. For a stricter rule, an immutable retention class is available on request, and that lock can't be lifted.

That's the RTO, and it depends on how many servers there are, the start-up order and the network settings. Nesnet doesn't state an RTO on this page; the time your environment needs is worked out with you and seen in a test bring-up.

The server keeps running for the whole backup. There is only a brief pause, from under a second to a few seconds, at the moment the snapshot is taken.

Yes, you can. Copies are brought up in an isolated network for testing without disturbing ongoing replication, and you check that your applications start. When the test ends, the test servers are removed.