The docs tree is laid out like the menus - a menu path is a docs path. Admin Panel / Settings follows the new structure: Visibility.md is Appearance.md, with new Features.md, Feature-Profiles.md, Security.md, Activities-and-Notifications.md and Delete.md; the section indexes, Admin-Panel.md, Admin-Panel-URLs.md (with the moved-pane redirects) and Validation.md follow the live menus, and Who-Sees-What.md explains in one place which layer decides what is shown and what is allowed. Pages are added for the menu entries that had none, each written from the code: every Admin Panel / Problems pane, the member menu, the board header, Board Settings, the All Boards left menu, and the list, swimlane and card menus. Older links to the moved panes are updated across docs/. TODO.md records what documenting the menus found: Edit Profile does not save a username for non-admins, the Support page's signed-out case is unreachable, Create board from this card is hidden from members who can edit the card, All Boards Settings is unreachable, and the Share template targets live in browser storage. Thanks to xet7 !
4.5 KiB
Admin Panel / Problems / Summary
The first pane of Admin Panel / Problems, and the one it opens on: what is
running now, what is wrong now, and how many new problems each problem stream
has. Address: /admin/problems/summary (the bare /admin/problems and the old
/admin-reports open it too). Site administrators only: the page shows "not
authorized" to anybody else, and every method behind it refuses a caller who is
not isAdmin. Organization administrators do not see it.
The pane reloads itself every 30 seconds. The same overview is printed by
snap run wekan.problems (Snap.md) and returned by
GET /api/admin/problems (REST API).
Status
From the systemStatusReport method (server/methods/systemStatus.js,
server/lib/systemStatus.js, models/lib/problemsOverview.js).
In progress - every migration or repair that is running now, read from the persisted status collections so it is right after a restart too:
- the database text migration (MongoDB to and from FerretDB v1), with its direction and phase (Migrations.md);
- the board data-repair, as
done/totalboards (Repairs.md); - data recovery / maintenance;
- an attachment migration or repair.
When nothing runs, the pane says "No migrations or repairs are in progress." While something does, it says logins can be slow until it finishes.
Problems - each detected problem as a line with its count and detail:
| Problem | Button | What the button does |
|---|---|---|
| Broken cards - cards with no board, swimlane or list, or an invalid type | Repair broken cards | Runs the repairBrokenCards method over every board, then says how many were fixed and how many cannot be placed on any board (no board id at all). |
| Lists missing their swimlane (#6670) | Restore | Runs restoreListSwimlanes: puts each list back in the swimlane it was created in, where that is still recorded and still exists, and says how many were restored and how many remain board-wide. |
| A failing login-page check | - | Configuration, not data; the detail says what to change (Login-Problems.md). |
| A failed or overdue Admin Panel backup | - | See Backup. |
| A data health finding | Repair | Runs repairDataHealth for that check. A repair that only moves data runs on one click; one that deletes data first turns into Confirm and runs on the second click (Data-Health.md). |
Every button is disabled while its repair runs, and the overview reloads when it is done so the count changes straight away.
Data health
The last run of the nightly read-only data health scan, as
<date> UTC - <checks that found something>/<checks>, or that it has never
run. Start runs the scan now (dataHealthScanNow); it runs in the
background and its findings appear under Problems above
(Data-Health.md).
Problems (the event streams)
One checkbox line per event stream with problems newer than its last
acknowledgement, written as the menu path
(Admin Panel → Problems → Security Report) with its count of new problems.
The streams are security, speed, tests, cpu, ram, disk, database,
integrity and mail (EVENT_STREAMS in models/eventLog.js); the API usage
report is not one of them. A row with severity info is never counted as a new
problem (models/lib/eventLogProblems.js), so a CPU spike that only records
its own mitigation does not raise the count.
Tick the streams you have reviewed, then:
- Acknowledge (
acknowledgeEventLog) - resets the new-problem count of each ticked stream. The rows stay in its pane. - Clear (
clearEventLog, #6893) - deletes every row of each ticked stream up to now and acknowledges it. A problem that happens again after the click shows as new again.
This is the only place problems are acknowledged or cleared; the stream panes themselves are read-only. With nothing new, the section says "No new problems."
Rows also age out by themselves after EVENTLOG_RETENTION_DAYS days (default
90, 0 keeps them for ever), checked once a day for the whole installation.
Related
- README.md - every Problems pane.
- Design.md - why each problem is one summary row.
- Security remediation design - the event log behind the streams.