Don't Migrate Off Opsgenie Until You've Paged Someone With the Replacement
Atlassian ended new Opsgenie sales on June 4, 2025. Support stops entirely on April 5, 2027. Somewhere between those two dates sits a shorter deadline that most migration checklists skip: the 120 days you get once you start Atlassian’s own migration tool. Use it, and Opsgenie switches off four months later, whether or not your escalation policies, on-call rotations, and integrations are actually working somewhere else.
That leaves a four-month window to prove a replacement works, not the nineteen months April 2027 suggests. Four months is enough time if you start with a working setup and confirm it before you touch the migration tool. It is not enough time if you start the clock first and figure things out as you go.
What Atlassian’s timeline says
Opsgenie’s retirement has moved in stages, and each one has already changed something for existing customers.
June 4, 2025: end of sale. No new purchases, no edition upgrades. Teams that grew since then are capped at the seats they already had.
October 2025: bundled access ended for anyone whose Opsgenie came packaged with Jira Service Management Cloud.
Now: every renewal signed in 2026 has to end before April 5, 2027, so it’s a short, terminal term no matter when you sign it.
Start plus 120 days: the automatic shutdown. Run Atlassian’s migration tool and Opsgenie turns off 120 days later.
April 5, 2027: login, the REST APIs, integrations, and the mobile app all stop working, for good.
The 120-day line is the one worth building a plan around. It’s the only deadline in that list that you control the start date of.
Atlassian sends you to two different places
Atlassian points ITSM teams toward Jira Service Management and DevOps teams toward Compass. Both handle alerting and on-call. Neither covers uptime monitoring, so whatever tool does that today keeps doing it, regardless of which one you pick. Teams trying to replace everything Opsgenie did sometimes end up paying for both.
Jira Service Management also bills per agent, and anyone who needs to receive a page has to be licensed as one. The number to plan around isn’t your service desk headcount. It’s everyone in the rotation, including the people who only get paged at 2 a.m. and never touch a ticket otherwise.
Prove it before you start the clock
You don’t have to choose between testing a replacement properly and keeping Opsgenie running while you do it. JAMS Incident Management’s trial runs for 14 days with every feature unlocked, 50 SMS and 10 voice calls included, and no credit card required. Point a copy of your alerts at it, rebuild your highest-severity escalation policy, and let both tools page the same person for a week.
A working test looks like this:
Start the trial and invite the whole rotation.
Rebuild one escalation policy and on-call schedule, then send a real alert through it.
Connect the source that matters most: Azure Monitor, ServiceNow, Zendesk, a webhook, or email over IMAP.
On day 14, decide. Nothing gets deleted automatically. Pick a paid plan, or stay on the free tier if your team is five people or fewer.
Only day 14 forces a decision, and even then, the decision is yours to make on your own timeline, not Atlassian’s.
What you’re replacing
Opsgenie’s job was routing alerts to the right person, escalating when nobody answered, and keeping the on-call schedule that decided who “the right person” was. JAMS Incident Management does the same three things through six channels: voice, SMS, push, email, Microsoft Teams, and Slack. Voice and SMS currently cover the United States, the United Kingdom, and Australia; the other four channels work anywhere.
It also does something a notification tool alone can’t: acknowledge a call and you can restart a failed job from the same call or text, without opening a laptop. Two-way integrations sync back to Azure Monitor, ServiceNow, and Zendesk, so acknowledging or resolving an incident in JAMS Incident Management updates the source system too. Single sign-on runs on OIDC, with setup guides for Microsoft Entra ID, Okta, Google Workspace, and Auth0. Everything is hosted on Microsoft Azure, with TLS in transit and provider credentials encrypted at rest. Teams already running JAMS Scheduler get one more thing for free: a failed job can open an incident automatically, without any extra setup.
Pricing runs per user, same as Jira Service Management, but published: free for teams of five or fewer, $6 per user a month for Starter (up to 25 users), $16 per user a month for Professional (up to 50 users, with SSO and premium integrations included), and custom pricing above that.
