Refresh the Backup and Restore guide with accurate information and best practices

[zmb3]

This page asks readers to back up the /var/lib/teleport/storage directory, but there is no storage subdirectory.

It should say

back up the /var/lib/teleport storage directory

[espadolini]

Backing up the directory while Teleport is running might lead to an inconsistent database state, so in addition to copying the files, we should probably also suggest sqlite-specific steps when appropriate, like sqlite3 /var/lib/teleport/proc/sqlite.db ".timeout 1000" ".backup proc.db" and sqlite3 /var/lib/teleport/backend/sqlite.db ".timeout 1000" ".backup backend.db".

[ptgott]

When would a user want to copy /var/lib/teleport? This guide has received only small changes in the last few years, but now that we're encouraging either (a) Teleport Enterprise (Cloud) use or (b) large, highly available self-hosted Enterprise clusters, would it make sense to rework the backup and restore guidance to:

1. Export audit logs and session recordings to a third-party platform
2. Manage configuration resources with git and an IaC tool (including, if necessary, tctl manifests in a directory)
3. Manage configuration files with git and whatever deployment tool you're using

The table of information to back up also recommends backing up certificate authorities—what would be the recommended way to do this?

[zmb3]

This was partially addressed by #50155