Tink vs restic
restic copies snapshots, forgets, and prunes. That is not a mechanic. Tink installs in one command, watches the Linux host, explains what is wrong, and helps you fix it.
restic runs the repo. Tink answers “why is this VPS sick?”
Accidental sysadmins land on restic because a snapshot repo looks cheaper than a NAS. Then the host volume fills, nginx dies, and restic check still reports no errors — no snapshot match for host fill. A filling /var is a Tink warning; it is not a restic job unless you scraped the rest-server and watched it.
Tink is the other job: detect the issue on the machine, say why it happened in plain English, propose the command, and run it only after you approve. Keep using restic if you already run snapshots. Use Tink if you run Linux servers and want a mechanic.
| Feature | Tink | restic |
|---|---|---|
| Setup time | 30 seconds (one curl | sh command) | Hours — RESTIC_REPOSITORY, RESTIC_PASSWORD, rest-server :8000, systemd timers, then a host that still has no mechanic after the last snapshot exits 0 |
| What you get | Working monitoring, diagnosis, and approved fixes | Deduped backups, snapshots, forget, and prune. Disk fill, nginx, Postgres, and certs stay your problem until you SSH past restic snapshots |
| Pricing | Free (Scout) / $9 / $29 per machine per month | BSD-2-Clause and free. You still pay the person who owns the VPS after the last snapshot stays listed |
| Hidden costs | None — fully managed | A quiet restic timer at 2am, a full repo that is not host fill, and a second tool to tell you the box is sick while restic check still reports no errors |
| Monitoring approach | Agent on the server — CPU, disk, services, logs, certs, ports | Snapshots, packs, and exit codes. restic does not watch a quiet cert or host auth.log after the last backup finished |
| Configuration | None after install — heuristics and AI | RESTIC_REPOSITORY, RESTIC_PASSWORD, exclude files, rest-server, rclone backends, and which systemd timer actually pages you |
| Plain-English diagnosis | Yes — AI explains root cause, impact, and fix | You read restic -v, then SSH into the host to change the Linux box restic already sits on |
| Fix execution | Proposes and executes approved commands with an audit trail | systemctl restart restic and restic unlock — restic cannot free host disk, rotate a cert, or stop SSH brute-force unless you wrote that job |
| Alerting | Built-in across 8 channels | Not a restic job — log lines plus a separate monitoring tool have to fire |
| Predictive alerts | Yes — disk fills in ~6 days, memory and CPU trends | Not a restic job — a snapshot ID does not forecast a quiet host filling up |
| SSH brute-force detection | Built-in — parses auth.log every scan | Never — restic does not read host auth.log, journald, or firewall state after the last snapshot landed |
| Machine offline detection | Agent presence monitoring with multi-channel alerts | Silence if restic dies after the last timer — unless you built a separate uptime check |
| Public status page | Shareable URL with 90-day history | None — restic has no customer-facing status page for the Linux host |
| Weekly fleet digest | Automated Monday digest + daily brief when issues are open | None — a snapshot recap is not a plain-English fleet narrative |
| On-call tracking | Built-in /oncall command + incident acknowledgment | Not included — wire failed restic into PagerDuty or another incident tool |
| Conversation interface | Telegram, WhatsApp, web dashboard, CLI | restic snapshots and SSH. No mechanic you text when /var filled and the last snapshot still showed added |
| Learning curve | None — works after install | Steep — repositories, rest-server, forget policies, prune, mount, and which failure is restic vs the running host |
| Best for | Freelancers, small teams, accidental sysadmins (1-50 Linux servers) | Teams that already run restic for backup snapshots, not a sick-VPS mechanic |
When restic is the right choice
Keep using restic when you already want that exact job:
- Backup snapshots with RESTIC_REPOSITORY — you need rest-server :8000, forget policies, and prune, not a mechanic.
- An existing backup stack — systemd timers, exclude files, and rclone backends matter more than guest nginx, certs, and SSH on a quiet VPS.
- An existing lab or nested setup — restic already starts, the repo already ships, and someone owns the restic box.
- Hosts that only page on backup or prune death — you already run restic logs, not a sick-box SSH loop.
The real cost of “just watch restic exit 0”
restic runs snapshots and prune. A working ops loop for five Linux boxes is not:
- A healthy snapshot still means you SSH into the host and change the box by hand when it gets sick later
- Host fill, nginx, certs, and SSH brute-force never appear unless a separate tool scraped those signals and a rule fired
- A filling /var is a Tink disk warning on every scan — restic will stay quiet until the next timer fails
- A single VPS with a dead agent still needs a mechanic, not another snapshot ID
For a 5-server team, Tink Mechanic at $45/month is cheaper than the engineer who keeps failed-job notifications alive after a free restic install, and every scan includes a diagnosis a quiet restic check will not type.
No RESTIC_REPOSITORY homework. No silent host. One command install.
Start MechanicAlso compare: Tink vs rclone · Tink vs rsync · Tink vs Syncthing