The only tool you need to know what is happening and how to fix it.
Go alertingexception-trackerissue-trackinglogsmetricsmonitoringobservability-platformopen-telemetryotelsession-replaytelemetrytraces
added X-Traceway-Synthetics header to monitors for analytics
narrow-adventure37fffd6eaa42
diff --git a/backend/app/synthetics/runner_http.go b/backend/app/synthetics/runner_http.goindex 07ef5c61..3989fbd1 100644--- a/backend/app/synthetics/runner_http.go+++ b/backend/app/synthetics/runner_http.go@@ -63,6 +63,7 @@ func probeHttp(ctx context.Context, check *models.SyntheticCheck) Outcome {if req.Header.Get("User-Agent") == "" {req.Header.Set("User-Agent", "Traceway-Synthetics/1.0")}+ req.Header.Set("X-Traceway-Synthetics", "1")transport := &http.Transport{TLSClientConfig: &tls.Config{InsecureSkipVerify: cfg.SkipTlsVerify},diff --git a/backend/app/synthetics/synthetics_sqlite_test.go b/backend/app/synthetics/synthetics_sqlite_test.goindex 10af37b1..0104a880 100644--- a/backend/app/synthetics/synthetics_sqlite_test.go+++ b/backend/app/synthetics/synthetics_sqlite_test.go@@ -205,7 +205,9 @@ func TestScheduleSkipsBrowserChecksWhenModeOff(t *testing.T) {func TestExecuteHttpCheckRecordsUp(t *testing.T) {projectId := setupSyntheticsDB(t)+ var syntheticsHeader atomic.Valueserver := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {+ syntheticsHeader.Store(r.Header.Get("X-Traceway-Synthetics"))w.WriteHeader(200)fmt.Fprint(w, "hello world")}))@@ -217,6 +219,10 @@ func TestExecuteHttpCheckRecordsUp(t *testing.T) {ScheduleOnce(context.Background(), now)ExecuteOnce(context.Background(), now)+ if got := syntheticsHeader.Load(); got != "1" {+ t.Errorf("X-Traceway-Synthetics = %v, want 1", got)+ }+if runs := allRuns(t); len(runs) != 0 {t.Fatalf("expected run row deleted after execution, got %d rows", len(runs))}diff --git a/docs/pages/learn/monitors.mdx b/docs/pages/learn/monitors.mdxindex 96ff1c04..5d9cbfdd 100644--- a/docs/pages/learn/monitors.mdx+++ b/docs/pages/learn/monitors.mdx@@ -23,6 +23,54 @@ Two properties matter operationally:- **A probe never runs late.** A queued run that outlives its window (the check's interval, capped at five minutes) is recorded as `missed`, not executed. A stale "is it up" answer would corrupt the uptime timeline.- **Downtime of Traceway itself shows as a gap, not as uptime.** If the backend was off, no runs existed, and the charts show unknown rather than a fabricated 100%.+## Keeping monitors out of your analytics++Monitor traffic is not real user traffic, so you usually want it out of Google Analytics and similar tools.++**HTTP checks** do not run JavaScript, so a client-side analytics tag never fires for them. Every HTTP probe also sends two headers you can match on:++```+User-Agent: Traceway-Synthetics/1.0+X-Traceway-Synthetics: 1+```++A custom `User-Agent` on the check replaces the default one. `X-Traceway-Synthetics` is always sent. Use it to skip server-side analytics, or to drop monitor requests from your access logs.++**Browser checks** run a real Chromium, so the page's analytics tag runs and every run counts as a visit. Traceway does not add the header to browser checks automatically. A header added to every request would also go to third-party APIs the page calls, and a cross-origin request carrying an unknown header fails its CORS preflight. That would break the page you are trying to monitor.++Instead, add the header in your script, scoped to your own origin:++```ts+import { test, expect } from '@playwright/test';++test('homepage loads', async ({ page }) => {+ await page.route('https://example.com/**', (route) =>+ route.continue({+ headers: { ...route.request().headers(), 'x-traceway-synthetics': '1' },+ })+ );+ await page.goto('https://example.com/');+ await expect(page).toHaveTitle(/Example/);+});+```++Your server can then leave the analytics snippet out of the page when the header is present. For example, in a server-rendered template:++```js+const isMonitor = req.headers['x-traceway-synthetics'] === '1';+// render the Google Analytics <script> only when !isMonitor+```++If you would rather not touch the server, block analytics requests from the script itself. The tag still loads, but nothing reaches Google:++```ts+await page.route(/google-analytics\.com|googletagmanager\.com|analytics\.google\.com/, (route) =>+ route.abort()+);+```++Do not rely on Google Analytics' built-in bot filtering. It works from a list of known bots and often lets headless Chrome through.+## AlertingAdd a notification rule of type **Monitor Down** on the Alerts page. It fires when any monitor in the project crosses its failure threshold, and sends a recovery notice when the check comes back. Pointing the rule at an escalation channel opens an on-call page instead; the page deduplicates per check, and recovery resolves it automatically, so a blip at night that recovers by itself never needs a human.