Opsgenie is going away. Atlassian stopped selling it on June 4, 2025, and support ends on April 5, 2027. If your team still gets paged through Opsgenie, you have a few months to choose a replacement and move over without a gap in coverage.
This post covers the dates, what Atlassian recommends, a catch in that recommendation that most migration guides skip, and a simple plan for a small team to switch on time.
The key dates
| Date | What happens |
|---|---|
| June 4, 2025 | Opsgenie end of sale. No new customers. |
| April 13, 2026 | Atlassian announces that Compass, one of the two recommended destinations, will be phased out in favor of DX. |
| April 5, 2027 | Opsgenie end of support. |
Treat April 5 as a hard wall, and aim to be fully switched over by early March. A migration that runs late tends to go wrong in the last week. That is when people find the forgotten integration or the schedule that only one person understood.
What Atlassian recommends
Atlassian's migration guide offers two paths:
- Jira Service Management for IT operations teams.
- Compass for developer alerting.
If your company already runs on Jira Service Management, moving there is the path of least resistance. Your schedules and escalations stay inside the Atlassian account you already pay for.
The catch: Compass is being phased out
On April 13, 2026, Atlassian announced that "Compass will be phased out" and encouraged Compass customers to move to DX. The announcement says there is no immediate change to access and promises early notice before anything changes, but it gives no end date.
The announcement does not mention Opsgenie, and the Opsgenie migration guide still lists Compass for developer alerting. So a team following the guide today could move its alerting into a product that is itself on the way out, and face a second migration later.
If you were planning on Compass, check Atlassian's current guidance first, or look at the other options below.
Start with what you actually use
Opsgenie can do a lot. Most small teams use a small part of it. Before comparing tools, list what yours really does:
- Schedules: how many rotations, how they hand off, and any overrides people rely on.
- Escalation policies: who gets notified, in what order, and after how long without an acknowledgement.
- Alert sources: which tools send alerts into Opsgenie, such as uptime monitors, Datadog, Prometheus, Grafana, or custom scripts.
- Notification methods: which people depend on phone calls or SMS to wake up, and who uses only the app or Slack.
- Heartbeats: any cron jobs or workers that ping Opsgenie to prove they ran.
- Configuration as code: Terraform or API scripts that create Opsgenie resources.
- Who touches it: the people who get paged, and anyone else who reads alerts or reports.
That list is your requirements document. It usually shows that the real question is narrower than "which incident platform should we buy?"
Your three realistic options
1. Move to Jira Service Management
This is best if you already pay for Jira Service Management, or your company standardizes on Atlassian. You keep one vendor and one login. The tradeoff is that you are buying into a bigger service management product, which can be more than a five-person team needs.
2. Move to a dedicated on-call and alerting tool
This is best if alerts come from many different tools, or people rely on phone calls and SMS to wake up. PagerDuty is the most common choice. As of November 2026, its pricing page lists:
- Free: up to 5 users, 1 on-call schedule, 1 escalation policy, and limited SMS and phone notifications.
- Professional: $21 per user per month billed annually, or $25 monthly.
Other tools in this space include incident.io, Rootly, ilert, and Better Stack. They differ mostly in pricing model, Slack workflow depth, and how much of the incident process they try to own. Compare total cost for your real team size, not the per-seat headline. We compare free plans, prices, and migration tools in Opsgenie alternatives in 2026, and costs for a team of 5 or 10 in cheaper PagerDuty alternatives.
3. Use a monitoring tool with on-call built in
This is best if most of your pages come from "the site or API is down." Many small teams use Opsgenie only to route alerts from their uptime monitor to whoever is on call. In that case you can drop a tool, not just swap one.
This is the option we built Tallwatch for. Each check runs from several regions, and an incident opens only when they agree. On-call schedules, escalation policies with timeouts, status pages, and heartbeat monitors for cron jobs come with it. The free plan includes 10 monitors, on-call, and escalation.
Where Tallwatch is not the right fit:
- It does not place phone calls or send SMS. Alerts go through email, Slack, Discord, Microsoft Teams, Telegram, and signed webhooks. If someone needs a phone call to wake up, route your last escalation level through a webhook into a voice service. We cover that in which alert channel wakes you at 3am.
- It is built around its own checks. If most of your Opsgenie alerts come from Datadog, Prometheus, or Grafana rules, option 2 fits better.
Which option fits your team
| Your situation | Best fit |
|---|---|
| Already paying for Jira Service Management | Jira Service Management |
| Alerts come from many tools, and people need phone calls | A dedicated on-call tool |
| Up to 5 people, one schedule, one escalation chain | A free on-call tier, such as PagerDuty Free |
| Pages are mostly "site or API is down" | A monitor with built-in on-call, such as Tallwatch |
| Planned to use Compass for developer alerting | Check Atlassian's current guidance first |
A month-by-month plan
December: inventory. Write down everything from the list above. Export schedules and escalation policies, or screenshot them. Ask each person on call how they get woken up today.
January: choose and set up. Pick a destination. Recreate one schedule and one escalation policy first, and connect your most important alert source.
February: run both side by side. Send alerts to both Opsgenie and the new tool for at least two weeks. Trigger a test incident on purpose, and make sure the right person is reached at every escalation level. Our guide on how to test your alerting has a checklist.
March: switch over. Make the new tool the one people acknowledge in. Remove Opsgenie from alert sources one by one, update runbooks and on-call docs, and cancel what you no longer need.
Before April 5: confirm nothing still points at Opsgenie. Search your code, Terraform, and monitoring configs for Opsgenie API keys and webhook URLs.
Mistakes to avoid
- Switching without overlap. Running both tools for two weeks is cheap insurance. A gap in paging is how outages go unnoticed for hours.
- Copying every old rule. Old setups collect alerts nobody acts on. Moving is a good time to drop them. See when to page vs notify.
- Forgetting heartbeats. Cron jobs that ping Opsgenie will go quiet when it shuts down, and a quiet heartbeat check looks just like a healthy one.
- Leaving it to the last week. The forgotten integration always turns up late. Give yourself a month of buffer.
The short version
Opsgenie support ends on April 5, 2027. Inventory what you use, choose a path that won't need another migration, and run both tools side by side before you switch. If your pages are mostly uptime alerts, you may not need a separate pager at all: Tallwatch's free plan includes multi-region monitoring, on-call schedules, and escalation policies in one place.



