Before you begin
- Atlas Flow on your plan (Atlas AVD) and turned on for the tenant
- An Atlas role with the Manage Atlas Flow settings permission (Atlas administrators have it)
Step 1: Create an automation rule
In Atlas Flow, open the Settings tab, find Automation and select Add automation rule. Give the rule a name, then choose when it opens a ticket:
- An Atlas alert fires: pick a condition such as Session Host Unhealthy or High CPU Alert, or choose Another alert type and enter an alert type. A type ending in a dot, such as agent.check., matches every alert in that family.
- An Atlas Reach device goes offline: optionally only after it has been offline for a number of minutes.
- An Atlas Reach check fails: optionally only at or above a severity, and only after it has been failing for a number of minutes.
- Provisioning or a deployment fails.
For alert and check rules, Lowest severity skips anything less serious. Only after (minutes) is useful for devices that drop off briefly, such as during a restart.
Step 2: Decide what the ticket gets
Under The new ticket gets, choose the ticket's priority (or leave it to follow the alert's severity), an assignee, a team and a category. A ticket that gets a team but no assignee is handed out by the team's auto-assignment, if the team uses it (see below).
Repeats, recovery and quiet hours
- While a ticket for the same problem is still open, the rule adds to that ticket instead of opening another one.
- Wait between tickets (minutes): after the rule opens a ticket, it waits at least this long before opening the next one.
- Resolve the ticket when the problem clears: on by default. When the alert or device recovers, and no other linked alert is still active, Atlas resolves the ticket and says so on its timeline.
- Don't open tickets during a maintenance window: set a start and end time, the days it applies to (leave them all unticked for every day) and its time zone. A window that ends before it starts runs overnight, such as 22:00 to 06:00.
Managing rules
Each rule in the list shows what fires it, how many tickets it has opened and when it last fired. Use the switch on a rule to turn it off without losing its settings, the edit button to change it, or the delete button to remove it. Deleting a rule leaves the tickets it already opened as they are.
Rules that open ConnectWise tickets are managed on the ConnectWise integration page. The Automation card counts them so you can see they exist.
Team auto-assignment
In the Settings tab, under Teams, each team has a New tickets setting:
- Leave unassigned: the default. New tickets for the team wait in its queue for someone to pick them up.
- Assign in turn (round robin): each new ticket goes to the next member in order.
- Assign to whoever has the fewest open tickets: the member with the lightest open workload on this tenant gets it. A tie goes to whoever is next in turn.
When auto-assignment applies
A team hands out a ticket when the ticket reaches the team with nobody assigned: a new ticket created for the team by a technician, by email intake or by an automation rule, or an unassigned ticket moved to the team. It never takes a ticket away from someone who already has it.
- Only team members who can work on this tenant are picked, so someone without access is skipped.
- Each tenant keeps its own turn order, and the fewest-open count looks at this tenant's open tickets.
- The ticket's timeline says it was assigned automatically and how, and the person who gets it is notified as usual.