5.5Health and testing
Last updated
Health states
Each connection carries a stored health reading:
| State | Meaning |
|---|---|
| Connected | The last bind succeeded. |
| Unreachable | The last bind failed. The reason is shown under the pill. |
| Not checked | No bind has been recorded yet. |
Under the pill, the list shows either the message from the last check or how long ago it was — "Checked just now", "Checked 12 min ago", "Checked 3 h ago", "Checked 2 d ago". The page header shows the age of the freshest reading on the page.
Background checks
You do not have to press Test to keep the readings current. The platform re-checks every connection in the background:
- the first check runs about 30 seconds after the platform starts,
- after that, every connection is checked every 5 minutes.
Each check is one bind per connection, no matter how many people have the page open.
An administrator of the server can tune this in the ActiveDirectoryHealthWorker section of the application settings:
| Setting | Default | Description |
|---|---|---|
ActiveDirectoryHealthWorker:Enabled | true | false stops the background checks; readings then change only when somebody tests |
ActiveDirectoryHealthWorker:Interval | 00:05:00 | How often every connection is checked. Values below one minute are raised to one minute |
ActiveDirectoryHealthWorker:StartupDelay | 00:00:30 | How long after startup the first check runs |
Being told when credentials stop working
A rejected bind password never heals on its own: the service account's password was changed in Active Directory, or the account was locked, disabled or expired. From then on every request through that connection fails until somebody enters the new password.
The platform can email you when this happens. The notification type is Directory Credentials Rejected ("Sent when a domain controller refuses the credentials stored for an Active Directory connection"). It is raised by any refused bind — a background check, a person's request or an integration's — and repeats at most every 12 hours for the same connection while the problem lasts. Once a bind succeeds again the alert clears, so a later failure alerts again. To choose who receives it, configure the notification rule; see Chapter 16, Notifications.
Last updated
Thank you for your feedback.