A tour of Utilities → Database backup, Restore and Backup schedule: every control and status

Know what every tile, badge, button and warning on the three Data protection screens means, so that a glance tells you whether last night's backup is safe and what to do when it is not.

Before you start

• You need to be a practice administrator. Open the sidebar's Admin group and choose Utilities. The three sections are in the Data protection group: Database backup, Restore and Backup schedule.
• This article is the screen-by-screen reference. For the setup itself, follow KB-067 — Set up nightly backups, a second copy and a monthly restore drill. To have a backup restored, follow KB-069 — Request a restore from backup.
• Two terms used below. The backup custodian is a separate program installed with the server that takes each new backup into a vault: a folder the server can read but cannot change. It exists so that ransomware running as the server cannot delete the backups.

The chip at the top of Utilities

Whichever Utilities section you are in, an administrator sees a chip beside the page title.

The last three are highlighted. Where a custodian keeps the backups, only a backup already in the vault counts as the last backup.

Database backup

The four tiles

Warnings above the history

Each of these appears only when it applies.

While a backup runs

A panel reads Backing up the practice database… with a progress bar and the current step: starting, running the dump, computing the checksum, then either copying to and verifying the second copy and pruning old backups, or "handed to the backup custodian". The practice can keep working while it runs.

Backup history

The card's heading line reads, for example, "12 on disk · pruned by the schedule, never below the three newest".

Back up now starts a backup. It reads Backing up… and is unavailable while one runs. Only one backup runs at a time.

Destination

Status

If there are no backups the card reads "No backups on disk yet. The first one is one click away…".

Verify a backup

1. Find the row in Backup history.
2. Choose Verify. Curaeon re-reads the whole file and recomputes its checksum.
3. Checksum verified. appears, and the badge becomes Verified.

If you see CHECKSUM MISMATCH — this backup is damaged and must not be restored from, the file on disk has changed since it was written. Take a new backup straight away and contact Support.

Download a backup

Use this for the off-site copy the practice must also keep.

1. Have encrypted storage ready: an encrypted drive, not a plain USB stick or an unencrypted laptop.
2. Choose Download on the row. The file is named with its kind, date and time and ends in .dump.
3. Save it straight to the encrypted storage.

The download is recorded in the Audit log as Backup downloaded before anything is sent, and it is listed in that week's Security review (KB-064 — Sign off the weekly Security review, and act on alerts and audit-log check warnings). If it cannot be recorded, it is not sent.

The note at the bottom

It repeats what Verify does and that downloads are recorded, and gives the address of the standalone console the server itself serves. Bookmark that address (KB-068 — Use the standalone admin console when the main app won't load).

Restore

This screen has no button that restores anything, on purpose.

Choose a backup

A table of every backup, most recent first, with two columns: Backup (date and size) and Restore with (the exact command for that file). The command box scrolls sideways when the command is longer than the column. The card does not appear when there are no backups.

The command names the backup file, and leaves the target database as a placeholder to be filled in at the server. That is deliberate: the command refuses to guess which database to replace.

Why restore is not a button

The second card gives three reasons.

Important: Do not run the printed command yourself. A restore is requested through Support with two approvals, and is run by a Curaeon engineer with whoever looks after your server (KB-069 — Request a restore from backup). This screen is where you find which backup you want and read its date to us.

A completed restore is recorded in the Audit log as Database restored from a backup, and the audit-log check reports it the next night (KB-069 — Request a restore from backup).

Backup schedule

Every change here saves by itself and shows Schedule saved — applies from the next run. There is no Save button. Each change is recorded in the Audit log as Backup schedule changed, with the old and new values.

Automatic backups

The run happens once a day at that minute. The server must be on at that time; a night it is off is a night with no backup, and the overdue warning follows after 36 hours.

Older backups are removed, "but never below the three newest, whatever their age". Where a custodian keeps the backups it does the removing, and it will not remove one younger than the minimum set when it was installed.

Destinations

Every backup lands on the server. This card is about the second copy and the alert.

Second copy directory shows one of two things:

• A text box, where the server keeps its own backups. Type the folder the server can reach, such as a mounted NAS share, and leave the field to save. Every backup is then copied there and the copy checked. As the screen says, "A copy on the same disk is a copy, not a backup."
• An explanation with no box, where a custodian keeps the backups. The second copy is "made by the backup custodian, from its own configuration", so it is set on the server by whoever runs it, not here.

Dashboard alert, when ticked, puts Backups need attention on the practice manager's dashboard whenever the last backup failed or none has landed in over 36 hours. It starts ticked. Leave it on.

The note at the bottom: there is no cloud copy, because clinical data does not leave the practice. For off-site protection, point the second copy at a drive you rotate off-site, and encrypt it.

Change the schedule

1. Open Utilities → Backup schedule.
2. Tick Nightly backup if it is off.
3. Set Run at to a time outside clinic hours, then click elsewhere to save.
4. Choose Keep backups for.
5. Check Dashboard alert is ticked.
6. Go to Database backup and confirm Next scheduled shows the new time.

What is recorded in the Audit log

Choosing Verify is not recorded.

If that didn't work

The screens do not show free space on the second location, the custodian's own log, or the date of the last restore drill. Ask whoever runs your server for those.

Still stuck? Raise a ticket at support.curaeon.com.au or call 1300 XXX XXX. If your clinic can't see patients right now, call and choose option 1. Support is staffed Monday to Friday, 8:00–18:00 Sydney time; outside those hours a call or text to the same number is answered on a best-effort basis.

Related articles

• KB-133 — Find your way around Utilities: every section, what it is for and who can open it — Find your way around Utilities: every section, what it is for and who can open it
• KB-067 — Set up nightly backups, a second copy and a monthly restore drill — Set up nightly backups, a second copy and a monthly restore drill
• KB-069 — Request a restore from backup — Request a restore from backup
• KB-068 — Use the standalone admin console when the main app won't load — Use the standalone admin console when the main app won't load
• KB-064 — Sign off the weekly Security review, and act on alerts and audit-log check warnings — Sign off the weekly Security review, and act on alerts and audit-log check warnings
• KB-135 — Export the patient list, billing history or the whole database from Utilities → Import & export — Export the patient list, billing history or the whole database from Utilities → Import & export
• KB-138 — Read Utilities → Storage and Clinical datasets: how big the database is, and which reference data is loaded — Read Utilities → Storage and Clinical datasets: how big the database is, and which reference data is loaded