Monitoring your workspace
A monitor is a page outside Depositori that watches several workspaces at once, such as In2Intel’s fleet page. It reads whether your workspace is up and how its background work goes, about once a minute. Only a workspace admin can let one in, and any admin can stop it again.
What a monitor reads, and what not
Section titled “What a monitor reads, and what not”A monitor gets a token that reads one thing: your workspace’s status.
- It reads: whether the workspace is up, which release it runs, whether its database, sign-in and site data are fine, and how its background jobs go (how many ran, failed or got stuck, per kind of job).
- It doesn’t read: any organisation, model, glossary, person or setting, and no error texts. It can’t change anything.
Let a monitor in
Section titled “Let a monitor in”The monitor starts it. It sends you to a page of your workspace that says Let its name monitor this workspace?, with what it will and won’t read and where the token goes.
- Sign in, if you aren’t yet: you come back to the same page.
- Check the name and the address it goes to.
- Choose Allow. Or Deny: then nothing is stored.
You go back to the monitor, which collects its token itself. Nobody copies a token. If you are not a workspace admin, the page says so and has no Allow: ask an admin to open the link.
See who watches, and revoke one
Section titled “See who watches, and revoke one”Under Workspace settings, Monitoring lists every monitor: its name, the address that received its token, when and by whom it was connected, and when it last read the status.
To stop one, choose Revoke and confirm. It stops at once. To connect it again, an admin allows it again from the monitor.
An agent that acts for a workspace admin can do the same with the list_monitors and revoke_monitor tools (see
For applications).