Nestview

Lightweight Docker Observability for Homelabbers

Nestview gives you a live health dashboard for all your Docker containers. Zero config, single container, no Grafana required. Drop it in your Unraid stack and get visibility into everything running on your server in under a minute.

---

Features

• Zero-config autodiscovery: detects all running containers automatically via the Docker socket

• Live health dashboard: CPU%, memory, uptime, restart count, and status; containers grouped by Compose project

• Network I/O stats: per-container bytes in/out with chart history

• Searchable log history: logs stored in SQLite and searchable from the UI

• Discord alerting: notifications on container crash, OOM kill, unexpected restart, and image updates

• Image update detection: checks Docker Hub and GHCR for newer digests; alerts you when updates are available

• Container actions: start, stop, restart, or pull & restart any container or Compose stack from the dashboard

• Authentication: mandatory login on first run with bcrypt-hashed credentials

• Multi-arch: amd64 and arm64 (runs natively on your Unraid hardware)

---

Getting Started

Install from Community Applications or pull the image directly: ghcr.io/kylejschultz/nestview:latest

Default port is 8484. On first launch, a setup wizard will walk you through creating your admin account.

---

Links

• Documentation and full setup guide

• Discord (fastest support)

---

Questions? Issues? Feature Ideas?

Drop a comment below or join the Discord for more direct assistance.

Edited May 14May 14 by kylejschultz

Changelog

Full changelog: https://github.com/kylejschultz/nestview/blob/main/CHANGELOG.md

---

v1.4.0 - 'Update & Restart' Process Improvement

• Update & Restart now performs real image pulls and container recreation, with durable operation status tracking.

• Container update checks now compare registry tags against the image each container is actually running, instead of only checking the mutable local tag.

• Recreated containers now keep you on the correct detail page without a temporary "container not found" state.

• Duplicate recreate events from stale collector snapshots are ignored.

• Concurrent update requests now return a clean conflict instead of racing overlapping operations.

v1.3.5 - Release Confidence Checks

• Release validation now runs in GitHub Actions for backend tests, frontend production builds, and Docker image smoke checks.

• Backend test coverage expanded around update-and-restart decisions and migration idempotency.

v1.3.4 - Log Streaming & Network Counter Fixes

• Log streaming now resumes from the latest persisted or buffered log timestamp instead of collector startup time.

• First-time log collection now backfills from the container start time, bounded by the configured log retention window.

• Docker log timestamps with nanosecond precision are now parsed correctly.

• Network counter resets are handled explicitly so reset counters become the new baseline instead of producing bad deltas.

v1.3.3 - Persistence & Network History Fixes

• Network history persistence now uses cached Docker counters so Nestview restarts preserve sane network deltas.

• Network history now resets cleanly after a watched container restart instead of carrying stale counters forward.

• SQLite now enables WAL mode, a busy timeout, and foreign key enforcement for better concurrent database behavior.

• Analytics daily ping state is persisted in the database so restarts do not send duplicate same-day telemetry.

• Main-line releases now post a structured Discord release card through the release webhook workflow.

• Removed the scheduled Discord repo stats channel updater workflow.

v1.3.2 - Bug Fixes

• Settings controls replaced with preset dropdowns (Log Retention, Metrics History) and a stepper (Session Expiry)

• Dark mode styling applied to the update check time picker.

• Network I/O charts fixed to show per-interval rates instead of cumulative totals.

v.1.3.1 - Disabling Swagger & OpenAPI endpoints

• Disabled the built-in Swagger UI (/docs), ReDoc (/redoc), and OpenAPI schema (/openapi.json) endpoints.

  • These were served unauthenticated by default and are not needed in production.

  • This reduces the exposed API surface for users running Nestview behind Cloudflare or a reverse proxy.

v1.3.0 - UI refresh and bug fixes

• Notifications page redesigned with global defaults and per-container exceptions

• Settings General tab redesigned with card grid layout and unified save bar

• New Settings > About tab showing version, uptime, database size, and Docker status

• Container detail page restructured - stats strip, side-by-side charts, paginated events

• Dashboard stack actions moved to a dropdown menu; exited containers visually dimmed

• Network history now preserved across container force-recreate

• Discord community link added to navbar

• Unmatched API paths now return 404 correctly

v1.2.1

Security hardening pass + a small quality-of-life addition for Discord alerting setup.

• Rate limiting applied to container action endpoints

• Constant-time comparison on login to prevent username enumeration via timing

• Password change now invalidates all active sessions immediately; other open sessions redirect to login automatically

• SSRF protection added to the image update checker

• Internal fields stripped from API responses

• "How do I get this?" help added to the Discord webhook URL field in the wizard and Settings - click it for a gif walkthrough showing exactly how to create a webhook. This is also available within the settings.

v1.2.0 - CPU/Memory History, Container Re-association, Security Hardening

• CPU and memory usage history charts on the container detail page - trends over your configured retention window

• Container history re-association - when a container is recreated with a new Docker ID, all logs, events, and history are automatically migrated instead of starting fresh

• Analytics "Learn more" modal with full disclosure of every field collected

• Analytics ping expanded with OS, container count, and release channel

• Sensitive settings values (webhook URL, password hash, session key) scrubbed from the settings API

• Clean container shutdowns no longer trigger crash alerts

v1.1.0 - Opt-in Anonymous Telemetry and Port Normalization

• Opt-in anonymous telemetry - a small optional daily ping to help track active installs. Off by default, you'll be prompted after setup/upgrade.

• Breaking change: The internal container port has been aligned from 8080 to 8484.

  • The CA template has been updated to reflect this. If the port doesn't update correctly after pulling the new image, remove and reinstall from the CA store.

v1.0.2 - Hotfix

• Fixed startup crash when optional env vars (LOG_RETENTION_DAYS, POLL_INTERVAL) are left blank in the Unraid CA template

v1.0.0 - Initial Stable Release

• Zero-config container autodiscovery via Docker socket

• Live health dashboard with CPU%, memory, uptime, and restart count; containers grouped by Compose project

• Network I/O stats with chart history

• Searchable log history with configurable retention

• Container actions: start, stop, restart, pull and restart per container or entire Compose stack

• Image update detection for Docker Hub and GHCR with Discord alerts

• Discord alerting on container crash, OOM kill, and unexpected restart

• Authentication with first-run setup wizard and bcrypt-hashed credentials

• Single container, single port, SQLite storage, no external dependencies

Edited Thursday at 06:47 PM2 days by kylejschultz

There seems to be an error in the template: on my system it defaulted to 8484:8484, but it the container listens on 8080...

13 hours ago, Krakout said:

There seems to be an error in the template: on my system it defaulted to 8484:8484, but it the container listens on 8080...

Yep, caught that once the template went live. I updated the template last night to use container port 8080/external port 8484 as it should be and it looks to be updated in the CA store. I just did a fresh install with the updated template.

Another change I just pushed to the template is ensuring that the Appdata path is pointing to the correct path inside of the container, as that was off as well causing a reinstall to be a fresh start rather than persisting data as expected. That change should be live shortly.

Give it another shot and let me know if you still run into issues with it.

Edited May 5May 5 by kylejschultz

v1.1.0 Update Note: Important Changes

Just wanted to put a note here to mention that with v1.1.0, I went ahead and normalized the port configuration so that both internal and external ports are configured to listen on 8484, instead of 8080 being the internal port. This means when you update, you will fail to reach your Nestview instance without updating your port configuration to reflect 8484:8484.

The CA template has already been updated to reflect this, so if you're a new user, you should be good to go. This will mostly just affect existing installations when upgrading to v1.1.0 or newer.