Loading…
Loading documentation...
Create point-in-time server snapshots, schedule off-host encrypted backups and verify restores on Servers Are Up without losing data.
Commands checked on 2026-11-12 with Servers Are Up Platform v1.7, sau CLI 0.2.0, restic 0.17.
You will know the exact operational difference between point-in-time hypervisor snapshots (for instant rollback before kernel or application upgrades), encrypted off-host backups (free to your own S3/R2/B2 storage or via catalog backup policies), and immutable WORM Recovery Vaults (/dashboard/recovery-vault), and you will have taken and verified a pre-change snapshot of a live server.
/dashboard/instances).snapshots:read, snapshots:write, and instances:read scopes if you want to trigger snapshots from CI/CD or the sau CLI.KVM snapshots capture crash-consistent block state instantaneously, and modern journaling filesystems (ext4, xfs, NTFS) plus PostgreSQL WAL recover cleanly. For maximum safety on busy databases, run a CHECKPOINT and filesystem sync right before snapping:
ssh -i ~/.ssh/id_sau_ed25519 -p 31001 ubuntu@203.0.113.25 \
"sudo -u postgres psql -c 'CHECKPOINT;' && sync"
In the web console, open Compute > Servers (/dashboard/instances), select your server, open its Snapshots / Backups tab, and click Create snapshot.
From a terminal or deployment pipeline, use sau snapshots create (which calls POST /api/v1/instances/{id}/snapshots and queues a durable provider command returning HTTP 202 Accepted):
sau snapshots create YOUR_INSTANCE_ID \
--name "pre-node24-upgrade-2026-11-12" \
--description "Clean baseline before upgrading Node.js and running Prisma migrations"
sau snapshots list YOUR_INSTANCE_ID
Wait until the snapshot status transitions from PENDING / CREATING to AVAILABLE before starting your risky change.
Snapshots live near the compute cluster for fast rollback; true 3-2-1 disaster recovery also stores encrypted backups in an independent storage fault domain. You have three built-in options:
/dashboard/storage), connect your own Cloudflare R2, Backblaze B2, Amazon S3, DigitalOcean Spaces, Wasabi, or SFTP target to back up Linux /mnt/data volumes with client-side encryption at $0 platform software fee.sau backups sources
sau backups quote YOUR_LINEAGE_ID --product backup-standard
sau backups create YOUR_INSTANCE_ID --notes "Pre-release recovery point"
/dashboard/recovery-vault) (25 GiB included free) with ransomware immutability locks and automated restore drills.If an OS upgrade or migration breaks your Linux plan server:
/dashboard/instances), select the server, locate your pre-node24-upgrade-2026-11-12 snapshot, and click Restore (or call POST /api/v1/instances/{id}/snapshots/{snapshotId}/restore with snapshots:write scope, which queues the restore command with HTTP 202 Accepted).--permanent, open Compute > Recycle bin (/dashboard/recycle-bin) before its retention window expires and click Recover.Confirm your snapshot is listed as AVAILABLE over the API and verify that a database dump or file backup can actually be read back:
curl -fsSL "https://serversareup.com/api/v1/instances/YOUR_INSTANCE_ID/snapshots" \
-H "Authorization: Bearer sau_your_token_here" | jq '.snapshots[] | {id, name, status, createdAt}'
An untested backup is only a wish—periodically restore a database backup onto a scratch container or trigger a verified restore drill on Operations > Backup & recovery vault (/dashboard/recovery-vault).
pg_dump / WAL archiving) in off-host object storage so you can recover code with a snapshot without rolling back customer rows.POST .../snapshots returns 403 Forbidden. Confirm your Personal Access Token includes snapshots:write (not just instances:write) and that no ReadOnly lock is attached to the resource in /dashboard/resource-manager.root. Run pg_dumpall | gzip as the postgres OS user in /etc/cron.d/ (0 2 * * * postgres /usr/local/bin/backup-db.sh) rather than root.sau snapshots create as the very first step of every production infrastructure upgrade script, and prune obsolete pre-upgrade snapshots after 7 days of stable runtime./dashboard/recovery-vault) so even a compromised admin token cannot purge recovery points./dashboard/key-vault).Every Servers Are Up virtual machine and container includes an integrated HTML5 browser-based SSH, RDP, and VNC console in your dashboard, default-off inbound firewall rules with per-IP allow-lists, and full programmatic control via REST API v1 and Model Context Protocol (MCP). Explore our Virtual Datacenter Plans or reach out to Servers Are Up Support.