-
Notifications
You must be signed in to change notification settings - Fork 1
Roadmap
What Watchtower does today and what is planned next — in plain terms. Everything runs on your server; no cloud account required.
Releases ship when ready (no fixed dates). Latest: 1.0.0 · Downloads and Releases · Changelog
Platform today: NeoForge 1.21.x. Fabric and older NeoForge lines are planned later.
First public release — the complete ops toolkit:
- Live dashboard (
:8787) — Overview, Live, Insights, Issues, Crashes, Mods, Backups, Activity, Session, Spark, Sources - Docs tab — bundled operator wiki with search
- Setup wizard — first-run audit, backup discovery, schedule, optional 2FA
- Scheduled health reports (default twice daily) with report retention
- Always-on background scan (~60s) for logs, crashes, mod errors, activity, lag spikes
- Performance history and Insights tab (busy/quiet hours, heatmaps, CSV export)
- Spark profiler integration — Spark tab with on-demand profile parsing
- External backup heartbeat for panel/cloud backups (Settings → Backups wizard)
- Login + optional 2FA, version chip, update banner
- Hosted-panel metrics honesty (cgroup labels, trust badges)
- DR CLI + DR Viewer
- NeoForge 1.21.x —
watchtower-neoforge-1.0.0+mc1.21.jar· GPL-3.0-or-later
Installation · Dashboard Overview · Disaster Recovery
Watchtower uses three update kinds — you do not need a full health report every time you open the dashboard.
| Kind | Plain English | When |
|---|---|---|
| Live | Charts while you watch | Every few seconds while dashboard is open |
| Background scan | Logs, crashes, recent activity | About once a minute while server runs |
| Full report | Complete fix list and deep mod check | When you run a report or on a schedule |
Open the Sources tab to see when each layer last updated. See Understanding-Data-Sources.
Still needs a full report: complete Issues queue, deep Mods analysis, Session playtime stats, report history.
Works without a fresh report: Live tab, crash folder scan, Activity ledger, lag investigations, mod log scan, backup rescan.
- Stop flagging Create (and similar server mods) as removable
- Optional Modrinth lookup for ambiguous jars — off by default
- Dependency tree — see why a mod is protected
- Fewer “unknown” crashes — real NeoForge stack attribution
- Watchdog stalls explained — pregen, contraptions, map render
- DR Fix tab matches the crash across dashboard, CLI, and DR viewer
- Startup boot profile — boot time, slowest phase, warnings trend
- Group identical crashes into one row
- Mark all reviewed in one click
- Notification bell on Overview (precursor to Discord alerts in 1.2.0)
Internal only — shared mod module for multiple Minecraft JARs. No operator-facing changes.
Names boot-time failures (mixin, loader, ecosystem patterns) as facts and brief text — mixin owner, conflicting mixins, duplicate jars, corrupt TOML, Create/Epic Fight/AzureLib fingerprints, and more.
Find owning jar from stack trace, corrupt jar detection, broken serverconfig scan, MCreator/Fabric-in-jar badges.
YAML rule packs under config/watchtower/rules/, builtin corpus, issue suppressions, watchtower rules validate CLI.
One view across many servers — compare TPS, issues, and backups.
Fabric 1.21 mod JAR; optional Discord/webhook alerts for crashes, lag, stale backups.
Second mod JAR for 1.20.x packs — pick +mc1.20 or +mc1.21 at download.
Full player analytics — retention cohorts, GeoIP, global playtime leaderboards, whitelist/OP/kick/ban from the dashboard.
GPU / graphics driver crashes — client-side patterns don't apply to headless dedicated servers.
Watchtower does surface roster, report-window activity, and peak concurrent players for ops triage — who was online during lag or crashes — without replacing Plan or Pterodactyl. Server-wide busy/quiet hours live on the Insights tab in 1.0.0.
Watchtower · GitHub · Releases · Modrinth · Contributing · GPL-3.0-or-later
Start here
- Home
- Installation
- Quick Start Checklist
- Dashboard Overview
- Dashboard Tabs
- Understanding Data Sources
- Using Spark with Watchtower
- Live Charts
Day-to-day
- Commands
- Configuration
- Security and Access
- Hosting Panels
- Reading Metrics on Hosted Servers
- Health Reports
- Backups
- Scheduled Reports
When things go wrong
Reference