Skip to content

v1.4.0 — Paging Reliability, Administrator Health Center & Release Safety

Latest
Latest

Choose a tag to compare

View all tags
@Dushyant-rahangdale Dushyant-rahangdale released this 23 Aug 17:52
· 162 commits to main since this release
Immutable release. Only release title and notes can be modified.
v1.4.0
3de1aac
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

OpsKnight v1.4.0 makes paging reliability a flagship capability, introduces a consolidated Administrator Health Center, establishes a release-quality contract, and publishes stable images for both AMD64 and ARM64.

🚀 Added & Highlights

1. 📟 Paging Reliability

  • Correct policy timing: Escalation steps preserve their configured delays instead of paging fallback responders immediately.
  • Atomic orphan recovery: Interrupted escalations are reclaimed through a single atomic claim, preventing competing workers from processing the same work.
  • Duplicate-delivery protection: Genuine escalation fallback no longer sends duplicate service notifications.
  • Reliable multi-channel routing: Unconfigured primary providers correctly fall through to available channels, and provider failures now trip circuit breakers.

2. 🩺 Administrator Health Center

A single administrative view now reports supported operational signals for:

  • Database connectivity and migration state
  • Scheduler, worker, and escalation backlog health
  • Notification-provider and integration failures
  • Public URL and encryption configuration
  • Current version and available upgrades

The standalone performance dashboard has been retired. Misleading per-process memory warnings and operator-attested backup freshness were removed because they could not accurately represent a distributed or externally managed deployment.

3. ✅ Release-Quality Contract

Every stable release is now gated by automated checks covering:

  • Clean installation and database migrations
  • Upgrade from the previous stable release
  • Fail-closed migration behavior
  • PostgreSQL backup and restore
  • Event trigger, acknowledge, and resolve lifecycle
  • Escalation and notification delivery
  • Helm and Kustomize rendering
  • Documentation-to-code capability coverage
  • Tagged AMD64 and ARM64 release images

4. 📦 Multi-Architecture Deployment

  • Stable container images now support Linux AMD64 and Linux ARM64.
  • Release tags include 1.4.0, 1.4, 1, and latest.
  • Startup migrations fail closed with bounded recovery instead of allowing the application to start against an uncertain schema.
  • The continuously updated test image remains AMD64-only for faster feedback.

5. 🔐 Security & Platform Hardening

  • Serialized first-administrator creation and protected the last active administrator.
  • Prevented disabled users from being reactivated through OIDC linking.
  • Added explicit administrator-controlled OIDC account-link approval and revocation.
  • Strengthened session revocation, RBAC, IDOR protection, input validation, and structured-log redaction.
  • Hardened status pages, exports, webhooks, custom fields, notification routing, distributed jobs, and incident state transitions.
  • Updated Next.js, Auth.js, Nodemailer, and vulnerable transitive dependencies to patched versions.

6. 📚 Versioned v1.4 Documentation

  • Added a 15-minute Compose getting-started path while retaining the longer first-week guide.
  • Added production Docker, Kubernetes, Helm, backup, recovery, troubleshooting, and upgrade guidance.
  • Published only supported APIs and integrations, with automated link and capability-coverage checks.
  • Kept v1.3 documentation historical while moving v1.4-specific behavior into its own versioned documentation set.

Upgrading

Back up PostgreSQL before upgrading and pin the exact release tag in production.

Docker / Compose

docker pull ghcr.io/opsknight-labs/opsknight:1.4.0
docker compose down && docker compose up -d

Kubernetes / Helm

helm upgrade opsknight opsknight/opsknight --set image.tag=1.4.0
  1. Deploy — database migrations run automatically and fail closed if they cannot complete safely.
  2. Verify Administration → Health Center after startup.
  3. Trigger a test incident and confirm acknowledge, resolve, escalation, and notification delivery.
  4. Review provider and integration health before routing production alerts to the upgraded deployment.

Full notes in CHANGELOG.md · v1.4 Documentation · Upgrade & Rollback Guide

Assets 3
Loading