Snapshot, schedule, and restore your server in minutes
The Backup page is one screen — no tabs. Stat tiles at the top, a snapshot trend chart, GFS retention breakdown, storage quota, the auto-backup scheduler, capture scope and the restore-points table. Manual snapshots take seconds; restores are previewed before anything is touched.
Before you make a backup
Backup is a Pro feature. The whole page renders an upgrade prompt instead of cards if your server is not on the right tier or the module is off.
- Backup module enabled in Server Settings (or guild is on Custom-Host).
- Server is on Pro tier — the page header carries a PRO chip.
- You hold the backup.view capability (default: every member can see), backup.create / backup.restore for action buttons.
- Storage quota has free space. The page enforces 5 GB per server; restore points are pruned by retention before quota is hit.
Section 1 — Stat tiles, top of the page
Five tiles summarise your backup posture in numbers.
What each tile reports
- Total restore points — every snapshot you currently have, manual and automatic combined.
- Last backup — relative time since the most recent successful snapshot.
- Dedupe ratio — how much storage SYNTHET is saving by sharing unchanged objects across snapshots. 3.2× is healthy.
- Unique objects — how many distinct items (roles, channels, emojis) live across all your snapshots. Adds context to dedupe.
- Storage used — total bytes consumed against the 5 GB quota.
Section 2 — Snapshot trend chart
A small line chart under the tiles showing how many restore points existed each day for the last 30 days.
What it tells you
- A flat line means your retention is stable — keepDaily/keepWeekly/keepMonthly are doing their job.
- A staircase climb means automatic backups are running but retention is not pruning. Check Section 5 settings.
- A sudden drop means you (or an admin) bulk-deleted snapshots from the table. Audit log will confirm.
Section 3 — GFS tier breakdown
GFS = Grandfather/Father/Son. SYNTHET keeps three tiers of snapshots and prunes them on different schedules.
- daily
- Every successful auto-snapshot becomes a daily until the keepDaily count is hit. Default 7. Used for "yesterday I broke something" recovery.
- weekly
- Once per week the latest daily is promoted to a weekly. Default keepWeekly is 4. Used for "we redesigned a category last month" recovery.
- monthly
- Once per month the latest weekly is promoted to a monthly. Default keepMonthly is 12. Used for compliance and long-term audit.
Section 4 — Storage quota
A horizontal bar showing bytes used vs the 5 GB cap. Hovering reveals the breakdown by tier.
What to do when it gets full
- Open the restore-points table and delete pinned manual snapshots you no longer need (auto ones get pruned automatically).
- Lower keepDaily / keepWeekly / keepMonthly in Section 5 — pruning runs the next time auto-backup fires.
- Reduce capture scope (see Section 7) — turn off includeMessages if you have it on; it accounts for >80% of size when enabled.
- Upgrade to Custom-Host — the 5 GB cap goes away.
Section 5 — Auto-backup schedule
The card that turns on automatic snapshots and decides when they run.
Controls in this card
Master switch. When off, no auto-snapshots ever fire and the rest of the card is greyed out.
Example: On
Three quick presets: Daily-only (light), GFS standard (recommended), Aggressive (everything). Sets cron and retention together.
Example: GFS standard
Five-field cron expression in your guild timezone. Edit directly if you need a custom schedule.
Example: 0 3 * * *
Read-only timestamps showing when the scheduler last fired and when it next will.
Section 6 — Retention
Three numbers and one toggle. Together they decide which snapshots survive pruning.
| Setting | Type | Default | Description |
|---|---|---|---|
| keepDaily | number | 7 | How many daily snapshots to retain before promoting/discarding. |
| keepWeekly | number | 4 | How many weekly snapshots to retain. |
| keepMonthly | number | 12 | How many monthly snapshots to retain. |
| autoPrune | toggle | On | When off, retention numbers are advisory and SYNTHET keeps everything until you delete manually. Use only for forensics windows. |
Section 7 — Capture scope
A collapsible card listing what each snapshot captures. Open it once when you set up the page; you rarely need to revisit.
Always-on scopes
- Roles — including hierarchy and per-role permission overwrites.
- Channels — name, topic, type, position, slowmode.
- Categories — and which channels live under them.
- Emojis and stickers — including custom uploads.
- Webhooks — name and channel binding (tokens are NEVER captured for security).
- Automod rules — every Discord-side automod entry.
- Channel and category permission overwrites.
- Guild settings — name, icon, banner, locale, AFK timeout.
Optional scopes
If on, captures the most recent N messages per channel. Vastly increases snapshot size and transfer time.
Example: Off
When on, snapshots are encrypted with a server-specific key before being stored. Adds a few hundred milliseconds to capture and restore.
Example: On
Optional whitelist. Empty = capture every channel. Useful for partial backups of sensitive content.
Channels skipped during capture. Wins over includeChannels when both are set.
Roles skipped during capture. Useful when a third-party bot manages a role you do not want versioned.
Section 8 — Restore points table
The biggest card on the page. A paginated table of every snapshot SYNTHET holds for your server.
Columns
- Type — chip showing manual / daily / weekly / monthly.
- Created — absolute timestamp in your guild timezone.
- Size — bytes consumed by this snapshot's unique data after dedupe.
- Scope — small icons indicating which scopes were captured.
- Author — Discord user for manual; "Auto" for scheduled.
- Actions — Preview restore, Restore now, Delete.
Restore preview modal
- Side-by-side diff of current server vs the snapshot.
- Counts of items added / changed / removed by tier (roles, channels, emojis, etc.).
- A scope picker — uncheck any tier you do not want to restore.
- Two buttons: Cancel (close) and Restore (execute).
Common tasks
Set up auto-backup for the first time
Open DashboardServerBackup.
- Section 5 → toggle "Enabled" on.
- Click the "GFS standard" preset button — it fills cron 0 3 * * * and retention 7/4/12.
- Section 7 → leave includeMessages off unless you really need it.
- Section 7 → leave encryptAtRest on (default).
- Wait until 3am tomorrow, then refresh the page; a new restore point will appear in Section 8.
Take a manual snapshot before a big change
- Click the green Create Backup button at the top of the page.
- Optionally name it (e.g., "Before category cleanup").
- Confirm. The snapshot completes in a few seconds for most servers.
- It appears in Section 8 marked Manual — pinned, so retention will never auto-prune it.
- Run your big change. If it goes badly, restore from this exact pin.
Roll back a misconfiguration without losing recent work
- Section 8 → find the most recent good snapshot, click Preview restore.
- In the modal, uncheck any scope you want to keep current. Common: keep Roles current, restore only Channels.
- Click Restore. SYNTHET shows a progress indicator; do not navigate away.
- When done, refresh Discord. Verify the restored items are correct.
- If the result is wrong, restore again from a different snapshot — the operation is idempotent.