Resources and the resource page
Every card, hexagon, alert and row of the resource list leads to the same place: the resource page. It answers what happened to this resource, when, and against which limit. This article explains how to find a resource in the list, what the page header tells you, what each tab holds and how to read the charts.
The resource list
Resources in the main menu lists everything monitored in the entity chosen in the header and below it.
| Filter | Use |
|---|---|
| Search name or host | Part of a name or a host name |
| Type | One resource type, or All types |
| Status | Critical, Warning, Unknown, Silenced, Healthy; several at once |
| Tag rule | key:value pairs separated by spaces, for example env:prod role:sql. A resource must match every term; a key on its own, such as client, matches any value of that tag |
| Include discovered databases and linked servers | Adds the items that discovery found under SQL Server instances, which the list leaves out by default |
Each row shows the Name, Type, Entity, Tags, Open alerts and Last data. Every filter is part of the page address, so a filtered list can be bookmarked or shared; Clear filters removes them all. Entity admins also see Add resource here.
The page header
The header says where the resource lives and how it is now:
- Location: Resources, then the entity (point at it for the full path) and, for a discovered item such as a database, the resource it belongs to.
- The name, the state and since when it has been in that state, the type and the tags. A resource whose checks have not produced any result yet is Unknown.
- The criticality, from 1 (low) to 5 (critical), used for routing and for the security score.
- Last data: when anything last arrived from it.
- Credential: the login its checks use, marked inherited when it comes from a parent resource.
When discovery no longer finds an item, such as a database that was dropped, the page says Not found by the last discovery. The item is kept with its history until an administrator deletes it.
Entity admins also see Test connection, Discover (for SQL Server instances), Edit and Delete. Deleting a resource removes its checks, metrics and alerts from Excubius only; the monitored system is not touched.
Tabs
The open tab is part of the address, for example ?tab=checks, so a link opens it directly.
| Tab | Content |
|---|---|
| Overview | Charts of the key metrics of the type, a summary of its checks (how many are on and how many fail to run), and what discovery found under it: availability groups, databases, linked servers |
| Checks | Every check with its state, last value, effective warning and critical thresholds and where each comes from, interval and last run. Search checks and Only problems narrow the list |
| Alerts | The open alerts of the resource. See its alert history opens the inbox filtered to this resource, with resolved alerts included |
| Timeline | What happened to it, newest first: alerts opened, escalated, acknowledged, assigned, silenced, notified and resolved, and the start and end of every maintenance window that covered it |
| Configuration | Its attributes; its identity (type, entity, credential, last discovery, and whether it appears on the public status page); and the Recent jobs: connection tests and discoveries with their results |
Reading the Checks tab
| You see | Means |
|---|---|
| baseline, entity Production, this resource | Where the threshold in force comes from. See Thresholds and baselines |
| Critical pending or Warning pending | The condition holds, but not yet for the whole for duration of the rule. The alert opens if it keeps holding |
| From another source | A source such as New Relic or Azure Monitor feeds this check, not the Excubius collector |
| A greyed row marked Off | The check is switched off on this resource. It stays listed so nobody forgets it |
| changed next to the interval | The check runs on its own interval on this resource, not the catalog's |
Editors and above can switch a check on or off, change its interval from 10 seconds to 7 days, and set warning and critical thresholds for this resource only. A level left alone keeps its inherited value, and Back to inherited thresholds removes the resource's own values. Thresholds for a whole branch are set on the entity by an entity admin.
Charts
Choose the Time window above the charts: 1 h, 6 h (the default), 24 h or 7 days. One window applies to every chart of the page, and it is part of the address too.
- Each chart draws the effective warning and critical thresholds as dashed lines.
- Where data is missing the line has a gap. A gap is never a zero.
- Hover, touch or use the arrow keys on a chart to read every value at one moment.
- Show as table lists the same values for reading or copying.
The charts move with the clock while the page is open, so the last point is always recent.
Connection test and discovery
An entity admin runs these from the header; you see the outcome on the page and in Recent jobs.
- Test connection asks the collector to connect with the resource's credential, and reports the server, version and latency, which permissions the login has or misses, or why it failed (Unreachable, Sign-in refused, Timed out…) with a hint.
- Discover reads the availability groups, databases and linked servers of a SQL Server instance and adds or updates them as resources under it. Items that disappear are marked missing, never deleted.
Both run in the background. A job waits for the collector, usually less than 10 seconds; not picked up means no collector took it.