User Manual¶
This section is for people using Tracedown: setting up checks, reading what came back, and getting told when something breaks. If you are the one running the installation, start with Installation instead.
The short version¶
You organise work into workspaces and projects, and the thing that actually gets monitored is a service: one probe script, one schedule, and a policy for which agents run it. The script is written in Lace, and it decides both what to check and what is worth alerting on.
That last point surprises people, so it is worth saying up front: there is no alert-rule builder. Alerting is expressed inside the probe script, next to the assertion it relates to. See Notifications for why.
Read in this order¶
-
The hierarchy, what a probe run actually is, and the vocabulary the rest of these pages assume.
-
Where things live and why the nesting exists.
-
Creating a check: schedule, agents, queue policy, run-now.
-
Lace in the editor: assertions, chaining, and emitting alerts.
Then, as you need them¶
| Page | Read it when |
|---|---|
| Variables | You need credentials or a base URL in a script, or a value carried between runs. |
| Reading Results | A probe failed and you want to know why, or you want the numbers in Grafana. |
| Notifications | You want alerts to reach a person or a webhook. |
| Silences & Quiet Hours | The alerts are reaching you too often, or at 3am. |
| Maintenance Windows | You have planned downtime and don't want it recorded as an outage. |
| Users & Permissions | Someone else needs in, or needs to stop getting paged. |
| Your Account | Two-factor, sessions, password, personal settings. |