Run a practice rebuild before your server needs one

The worst time to learn what your backup is missing is the night your server dies. A rehearsal into a spare server turns a RestoreCord snapshot from a hope into a tested plan.

Short answer

Invite your RestoreCord bot to an empty spare server, restore a recent snapshot into it without clearing anything live, and time how long structure and history take. Check permissions and private channels, log what went wrong, and book the next drill.

Most owners take a snapshot, see the confirmation and never look at it again. Then the server is wiped, and they discover in the middle of the crisis that the bot role was too low, a private channel came back public, or the restore takes far longer than they promised their members. A drill moves all of those surprises to a quiet afternoon.

What a drill tells you

QuestionHow the drill answers it
Does the snapshot hold what I think?You see the restored channels, roles and messages with your own eyes
How long will a real rebuild take?You time structure and history separately on your own server size
Is my bot set up correctly?Missing permissions or a low role show up as failures, not in an emergency
Do my permissions survive a restore?You can open every private channel and check who sees it

Prepare the test ground

  1. Create an empty server that exists only for drills. Never use a server with real members in it.
  2. Invite the same RestoreCord bot your community verifies with. A bot has to be installed in a server before it can restore into it.
  3. Give it Manage Roles and Manage Channels and drag its role to the very top.
  4. In the dashboard, open Snapshots and preview the snapshot you plan to use. The previewer is on every plan, Free included.

Double-check the destination before you confirm. Choose the spare server, and leave clear destination off unless the spare server is the one being cleared. Clearing deletes existing structure.

Run it and keep the clock going

Start the restore and note the time. RestoreCord brings the structure back first and the message history after it, all at rates Discord accepts. That is why a server with many channels and a high message count takes longer: the pace is set by Discord, not by a queue you can skip. Let the restore finish and don't start a second one on top of it. The time you measure here is the number to give your staff when they ask how long a real recovery will take.

Inspect the result like a stranger

Walk through the rebuilt copy as if you were a new member, then as a moderator:

  • Every category and channel is present and in the right order
  • Staff and private channels are still hidden from @everyone
  • Roles exist with the permissions you expect, nothing more
  • Recent messages appear in the channels where history matters to you

A snapshot holds channels, roles, settings, member role relationships and messages up to the per-channel count you chose. It does not hold the members themselves, which is why a member pull is a separate step. In a drill, don't pull your community into the spare server; test the member side by verifying with a spare account instead.

Write it down

A drill nobody recorded is a drill you will have to repeat from memory. Keep a short log like this in your staff channel or notes:

Rebuild drill logtext
Drill date:          ____-__-__ (UTC)
Snapshot restored:   name + creation date
Pinned?              yes / no
Destination:         spare server name
Messages / channel:  ___ (plan limit: ___)
Restore started:     __:__ UTC
Structure finished:  __:__ UTC
History finished:    __:__ UTC
Problems found:      e.g. #staff visible to @everyone
Fixes applied:       what changed, new snapshot taken?
Next drill:          ____-__-__

Know the limits you are testing

PlanSnapshots per serverMessages per channel
Free150
Premium5200
Business30500
Enterprise502,000

Pin the snapshot you rehearsed with. On Business and Enterprise, scheduled snapshots rotate out the oldest unpinned one when you reach your limit, but pinned snapshots are never removed automatically.

When the real thing happens, follow rebuilding a deleted server and then restoring roles. Never took a snapshot? Start with rebuilding with no backup.

A rebuild drill in four steps

  1. Take or pick a snapshot

    Use a recent snapshot of the live server, ideally a pinned one so rotation never removes it.

  2. Prepare a spare server

    Create an empty server, invite the same bot and move its role to the top.

  3. Restore and time it

    Restore the snapshot into the spare server and note when structure and history finish.

  4. Inspect and write it down

    Check permissions, private channels and roles, log every issue and set the next drill date.

Signs your drill passed

Tick these off before you close the spare server for the day.

The bot restored without permission errors

Its role sat above everything it had to create.

You know your recovery time

Start and finish times for structure and history are in the log.

Private channels came back private

You checked them with a non-staff account.

The next drill is booked

A date is in the log, and the rehearsed snapshot is pinned.

Rebuild drill questions

Test the backup, then stop worrying about it

Take a RestoreCord snapshot, rehearse it once, and you will know exactly how your server comes back.