Grafana OnCall is shutting down — what Slack teams should use instead

Published August 8, 2026 · A practical migration guide for teams that self-hosted Grafana OnCall.

What Grafana Labs announced

Grafana OnCall OSS was a popular open-source on-call management tool that teams could self-host alongside their Grafana stacks. On March 11, 2025, Grafana Labs placed the project in maintenance mode — no new features, security fixes only. Then on March 24, 2026, the repository was archived and the Cloud Connection was deactivated.

That Cloud Connection was the critical piece: it was the relay that Grafana Labs' infrastructure provided to make phone calls, send SMS messages, and push notifications to the mobile app. Without it, self-hosted deployments lost all three of those notification channels simultaneously.

The official migration path is Grafana Cloud IRM, Grafana Labs' paid managed successor. It starts at $20 per active user per month. If you were self-hosting OnCall to control costs, this is a meaningful shift in how you'll be billed going forward.

Why Grafana Cloud IRM isn't automatically the right answer

Grafana Cloud IRM is the natural continuation if you're deeply embedded in the Grafana observability stack — if you already run Grafana Cloud for metrics, dashboards, and alerting rules, extending into Cloud IRM keeps everything in one place and your alert routing stays tight-coupled to your existing setup.

But for many teams, Grafana OnCall was chosen precisely because it was free to self-host. At $20/user/month with per-seat pricing, a 10-person on-call rotation costs $2,400/year before anyone touches a status page or an escalation chain. A 30-person team hits $7,200/year. That's a significant step up from $0.

The better question to ask first: what were you actually using Grafana OnCall for?

Be honest about what you actually used Grafana OnCall for

Grafana OnCall covered a wide feature surface, but most teams we've spoken to were primarily using one of three things:

  1. Rotation scheduling. Defining who is on-call, when, and for how long — weekly handoffs, daily rotations, follow-the-sun coverage, override management, and a live answer to "who is on-call right now?"
  2. Slack notifications. Pinging the on-call person in Slack when an alert fires, keeping a Slack user group like @oncall current, and sending shift handoff reminders.
  3. Phone and SMS escalation. Multi-tier escalation chains that call or text people who miss a page, mobile push notifications, and escalation policies that involve on-call managers or secondary responders.

If your usage is primarily (1) and (2), the deactivation of the Cloud Connection barely affects you — you weren't using phone or SMS alerting anyway. That means you don't need a product that replaces the full Grafana OnCall feature set. You need a rotation tool.

The realistic alternatives, ranked by team profile

If you're a Slack-first team that mainly needed rotations → Sched

Sched is a Slack-native rotation manager. It runs your weekly, daily, or custom rotations; keeps a Slack user group in sync with whoever is currently on-call; sends handoff messages; and lets people swap shifts directly from Slack.

  • Flat pricing: free for up to 3 schedules, $9.99 per workspace per month for unlimited.
  • No per-seat pricing — adding people to rotations never changes your bill.
  • Setup takes minutes; no server to run, no Kubernetes manifests to maintain.
  • Best fit for teams up to ~200 engineers running 1–50 rotations.

Trade-off: Sched is intentionally focused on rotations and Slack. If you relied on Grafana OnCall for phone/SMS escalation chains, you'll need a separate alerting tool alongside it — or one of the heavier platforms below.

If you're already embedded in Grafana Cloud → Grafana Cloud IRM

If your organization pays for Grafana Cloud for metrics and dashboards anyway, Cloud IRM is the path of least resistance. Your alert rules stay in the same system, your integrations carry over, and Grafana provides a migration path from OSS. Expect per-seat pricing starting at $20/active user/month on the Pro tier.

This makes sense if you genuinely need the tight coupling between your Grafana dashboards and your on-call routing. It does not make sense if you self-hosted OnCall specifically to avoid per-seat pricing.

If you need full incident management → PagerDuty or Incident.io

PagerDuty is the most established all-in-one replacement: rotation scheduling, escalation policies, phone and SMS paging, hundreds of alert source integrations, mobile app, status pages, and postmortems. Incident.io is a newer, more Slack-native alternative that bundles on-call with incident response and tends to feel more modern.

Both are significantly more expensive than self-hosted Grafana OnCall was. Make sure you actually need the incident management features before committing to the pricing.

If you want to keep self-hosting → GoAlert or OneUptime

GoAlert and OneUptime are open-source on-call tools you can self-host at zero licensing cost. The trade-off is real infrastructure overhead: you'll need to run a PostgreSQL cluster, Redis, and a message broker, plus wire up your own Twilio account for phone and SMS alerting. If your team has the ops appetite for it and genuinely needs to keep everything on-premises, this is the path. Most teams don't.

A simple decision tree

1. Did you mainly use Grafana OnCall to schedule rotations and notify people in Slack — without relying on phone or SMS alerts?

→ Yes: Move to a Slack-native rotation tool like Sched. You don't need to replace the full Grafana OnCall feature set.

2. Are you already paying for Grafana Cloud for metrics and dashboards?

→ Yes: Evaluate Grafana Cloud IRM. The integration is tight and the migration from OSS is supported. Compare the per-seat cost at your team's size before committing.

3. Do you need phone/SMS escalation, multi-tier escalation policies, or a large library of alert source integrations?

→ Yes: Look at PagerDuty or Incident.io. Compare total cost per active responder, not just list price.

4. Do you need everything on-premises and have the ops capacity to maintain it?

→ Yes: Evaluate GoAlert or OneUptime. Factor in the Twilio setup and infrastructure maintenance before deciding.

How to migrate cleanly

The migration steps are largely the same regardless of which tool you land on:

  1. Export your current rotations. Pull your Grafana OnCall schedule configuration before the repository becomes any harder to access. Note the cadence, participants, any overrides, and holiday coverage.
  2. Map your alert sources. List every system that routes alerts into Grafana OnCall today — Grafana alert rules, Prometheus Alertmanager, Datadog webhooks, custom integrations. These need to be re-pointed before you cut over.
  3. Run the new tool in parallel. Re-create your rotations in the new system and run both side-by-side for a sprint or two. Compare who got paged and when — discrepancies are easier to catch here than at 3am during an incident.
  4. Cut over deliberately. Switch alert routing on a Monday, not a Friday. Keep your old Grafana OnCall config readable for at least 30 days as a reference.
  5. Document the new setup. On-call configuration accumulates tribal knowledge. Write down what exists and where it lives — your future self will thank you the next time something changes.

FAQ

When exactly did Grafana OnCall shut down?

Grafana OnCall OSS entered maintenance mode on March 11, 2025. The Cloud Connection — which powered phone calls, SMS, and mobile push notifications for self-hosted deployments — was deactivated on March 24, 2026. The GitHub repository was archived the same day. The codebase remains available under AGPLv3 but is no longer actively maintained by Grafana Labs.

Can I still fork and run Grafana OnCall myself?

Technically yes — the code is still there. But you would need to build your own equivalent of the Cloud Connection to handle phone calls and SMS (typically via Twilio or a similar provider), and you'd be responsible for all ongoing security patches with no upstream maintenance. For most teams this is more work than switching to a maintained alternative.

Our team only used Grafana OnCall for rotations and Slack — what should we do?

The deactivation of the Cloud Connection most likely didn't affect you directly if you weren't using phone or SMS alerts. But you're now running unmaintained software, which is a security and reliability risk. Move to a maintained alternative. For Slack-first teams that just need rotations, a Slack-native tool like Sched covers exactly what you were using Grafana OnCall for.

Can Sched replace Grafana OnCall for our team?

If you used Grafana OnCall primarily to run rotations and surface the on-call person in Slack: yes. Sched runs the schedule, syncs your Slack user group, sends handoffs, and lets people swap shifts. If you relied heavily on phone/SMS escalation chains, you'll want either Grafana Cloud IRM, PagerDuty, or a separate alerting tool alongside Sched.

You can install Sched in your Slack workspace and try it free for up to 3 schedules — no credit card required.

Further reading