Operational health shows whether Advanced Boards: Kanban, Analytics & Forecasts keeps up on this site: the app version, the background work, the storage used and the recent jobs. It also holds one control, the pause on changes.
Open Operational health #
Open Administration, then select Operational health in the section list (Administration overview and license).
Read the Status card #
While the license is inactive or Atlassian has not confirmed it, a refusal stands in place of the Status card. Pause changes and Details for support stay.
| Row | What it says |
|---|---|
| App version | The version now running on the site |
| Update | “Still being applied”, shown only while a new version finishes its update by itself |
| Background work | “Up to date”, or how far behind the changes from Jira are, such as “25m behind”, shown only after about 20 minutes; Last pass failed in red, with an (i) that names the passes, or Late in amber when the five-minute sync is late; “Nothing imported yet” on a site with no board |
| Recent background jobs | The jobs by state, failed first and done last, such as “2 failed, 118 done”. The states are failed, queued, running, waiting to retry, paused for authorization, paused for the license, waiting for Atlassian to confirm the license, paused for the budget, stopping, canceled, replaced and done. Select the failed count, or Open Imports and jobs, to read each job. Its (i) “About the jobs counted here” says finished jobs are kept for up to two days, except the run that built each board’s current data. With none, it reads “Nothing is running, and no finished job is kept.” |
| Storage used | The app’s share of its storage, such as “35%, new exports stop first at 60%” |
| Audit trail | Shown only while a change is missing from the trail: “Catching up” in amber or “Incomplete” in red (Audit trail) |
The (i) beside Storage used says what is counted: board history, saved results, exports, alerts, settings and activity records, with room kept to rebuild the largest boards and an allowance for logs and settings. The row shows “Nearly full” in amber or “Full” in red. As storage fills, the app refuses new work in steps:
| Storage used | What the app refuses |
|---|---|
| 60% (“Nearly full”) | New exports |
| 65% | New saved results, such as forecast and scenario runs |
| 70% | A new board’s first import, or a board’s source widened |
| 80% (“Full”) | Board refreshes from Jira and new alert evidence |
Deletions, saves of settings and the audit trail are never refused.
To free room, choose a shorter history window (Data retention and deletion), delete saved results, or delete a board.
Pause changes #
Pause changes refuses changes and new forecast and analysis runs for everyone on the site, until an administrator lifts it. Canceling a job or an AI investigation, Import again, turning error reports off and deleting all app data still work. Reading and exporting keep working. The card shows Current state: “Changes are not paused” in plain text, or a yellow “Changes are paused” lozenge, with “Since” and the time. The (i) About pausing changes beside the card’s title says in one sentence what a pause refuses and what keeps working.
- Select Pause changes.
- Read the dialog “Pause changes for everyone?”: everyone is refused a change and a new run until an administrator lifts the pause.
- Note that work the app does by itself goes on (imports, refreshes and the daily cleanup), and so does work started before the pause, until you cancel it in Imports and jobs.
- Select Pause changes to confirm, or Keep changes allowed to leave.
To end the pause, select Allow changes again. The flags read “Changes are paused.” and “Changes are allowed again.” The button is held while a save is under way, while the pause is being read again, or if your administrator role could not be confirmed.
Scheduled passes #
Details for support at the page’s foot is a closed card: select Details for support to open it. Inside, Scheduled passes lists four passes: “Work item sync (5 minutes)”, “Reconciliation (hourly)”, “Retention sweep (daily)” and “Deep reconciliation (weekly)”. Below them, under Support and diagnostics, stand Run diagnostics and Build support bundle (Support and diagnostics). Each pass shows the time of its last run and how long ago it was, such as “(3m ago)”, with a state:
- On time: it ran within its period. This is plain text.
- Due: it is up to two periods old. This is plain text, not a problem.
- Late: it missed more than two periods. The lozenge is amber.
- Last pass failed: it ran and recorded a failure. The lozenge is red; where the app knows why, its (i) says it.
- Not run yet: it has not reported.
The work item sync runs every 5 minutes, and a board takes in changes from Jira at most every 15 minutes. So a change in Jira shows within about 20 minutes.
If something looks off #
- “Late” beside a pass: check again later, then contact support through Get Help.
- “Last pass failed”: the pass runs again by itself. The (i) “Which passes failed” names them.
- “Still being applied” that stays: contact support through Get Help.
- “The pause on changes was not saved.”: select the button again.
- “The app did not confirm whether the pause on changes was saved.”: reload the page and read Current state.
Who can open it #
Only a Jira administrator can open Administration and set or lift the pause.
