Settings
Synthetics settings are under Settings → Synthetics (or the gear on the Synthetics pages). Each tab shows only for roles with access to it.
| Tab | Description |
|---|---|
| Locations & agents | Locations, agents and registration tokens — see Locations & Agents |
| Maintenance windows | Planned muting and pausing — see Alerts & Maintenance |
| Notification channels ↗ | Opens the KubeSense notification channels that monitors alert to |
| Global variables | Values and secrets shared by monitors |
| Defaults & usage | Monitor defaults, limits, browser test integrations, tag policy and monthly usage |
Global Variables
Variables hold values that several monitors share — base URLs, API tokens, passwords — so they are changed in one place. Use them in any monitor field as {{ NAME }}, for example {{ BASE_URL }}/health or Authorization: Bearer {{ API_TOKEN }}.
Click New variable:
| Field | Description |
|---|---|
| Name | Upper case, digits and _, starting with a letter and at least 3 characters, e.g. API_TOKEN |
| Type | See below |
| Value | Up to 64 KiB |
| Description | What the variable is for |
| Tags | e.g. env:prod, for filtering |
| Type | Description |
|---|---|
| Plain | A visible value |
| Secure | A secret: write-only, encrypted at rest and masked as *** in results and exports |
| TOTP | A one-time-password generator for two-factor logins: a Base32 secret or an otpauth:// URI, digits, period and algorithm. {{ NAME }} gives the current code |
| From test | A value extracted by an HTTP monitor or a multistep step (for example a fresh token), refreshed each time that monitor's scheduled run passes. Run now and tests do not refresh it |
Open a variable to see where it is used: every monitor that references it, and in which field.
Built-ins and filters
Every field also accepts built-ins, computed per run:
| Built-in | Value |
|---|---|
{{ uuid }} | A random UUID |
{{ timestamp(-3600, s) }} | Unix time plus an offset, in s or ms |
{{ date(0d, YYYY-MM-DD) }} | The UTC date plus an offset (-1d, 2h…), in a format |
{{ numeric(8) }}, {{ alphabetic(8) }}, {{ alphanumeric(8) }} | Random characters |
{{ monitor-id }}, {{ result-id }} | The monitor and the run |
A reference can take one filter: {{ NAME | urlEncode }}, urlDecode, base64Encode or base64Decode. Saving a monitor checks every reference; an unknown variable is refused with the field it is in.
Defaults & Usage
Monitor defaults
Values a new monitor starts with, unless it sets its own:
| Default | Description |
|---|---|
| Frequency, Timeout | Schedule |
| Fast retries, Retry interval | Retries of a failed attempt |
| Locations | Locations picked for a new monitor, e.g. in-cluster |
| Labels | Labels added to a new monitor |
| Browser test device | Device of a new browser test |
| Failure threshold, Recovery threshold, Location threshold | Alert policy |
| Alert channels | Notification channels of a new monitor |
Limits
Limits for all users, each at least as strict as the installation's own (shown beside each field; see Configuration):
| Limit | Description |
|---|---|
| Max monitors | Most monitors that can exist |
| Minimum frequency | Shortest interval, overall and per check type |
| Max locations per monitor | Most locations one monitor may pick |
| Max steps per multistep test | Most steps in a multistep test |
| Max concurrent on-demand runs | Run now, tests and CI batches running at the same time |
Browser test integrations
| Setting | Description |
|---|---|
| APM trace URLs | URL patterns (with * wildcards) of your own services. Requests a browser test makes to them carry a trace header, so the run links to the backend traces in KubeSense. List only your own services: third parties receive the header too |
| Mark browser test sessions in RUM | Pages that load the KubeSense RUM SDK report browser test visits as synthetic sessions, not real users |
Tag policy
Rules for monitor labels (tags), so every monitor carries, for example, a team and an env:
- Each rule has a Tag key, optional Allowed values and an optional Pattern (regular expression).
- With Enforce the tag policy on, monitors that break a rule cannot be saved or imported. Off, the rules are advisory: violations are flagged in the monitor list and counted on the Overview.
Usage
Monitors and runs per month, with succeeded and failed runs per check type. Pick a month to see earlier ones.