fix(common-utils): keep metric series from different scopes apart

.changeset/metric-series-scope-name.md

@@ -0,0 +1,14 @@
+---
+'@hyperdx/common-utils': patch
+'@hyperdx/app': patch
+'@hyperdx/api': patch
+---
+
+fix: keep metric series from different instrumentation scopes apart
+
+Two instrumentation scopes can report the same metric with the same resource,
+scope and point attributes. Metric charts and alerts keyed a series on those
+attributes only, so they read the two scopes as one series: a gauge kept one
+scope's value, a counter's `increase` and a histogram's count compared one
+scope's total with the other's, and a group-by on `ScopeName` showed one scope.
+The series key now includes `ScopeName`.

packages/api/src/clickhouse/__tests__/renderChartConfig.int.test.ts

@@ -1139,7 +1139,7 @@ describe('renderChartConfig', () => {
       //   Row B: ResourceAttributes={service.name: api},           Attributes={host: h1}
       //
       // AttributesHash is now computed as the variadic
-      //   cityHash64(ScopeAttributes, ResourceAttributes, Attributes)
+      //   cityHash64(ScopeName, ScopeAttributes, ResourceAttributes, Attributes)
       // for all metric schemas (Map and JSON), so the two rows hash
       // distinctly: each lands in its own series. Both rows are then the
       // first row of their respective series, the lagInFrame Rate is NULL

packages/api/src/tasks/checkAlerts/__tests__/checkAlerts.int.test.ts

@@ -7738,7 +7738,7 @@ describe('checkAlerts', () => {
 
       // Insert gauge metrics for two different services
       // Note: ResourceAttributes must differ per service so that
-      // AttributesHash (variadic cityHash64(ScopeAttributes, ResourceAttributes, Attributes))
+      // AttributesHash (variadic cityHash64(ScopeName, ScopeAttributes, ResourceAttributes, Attributes))
       // produces distinct hashes. Otherwise, the Bucketed CTE collapses all rows into one group.
       const gaugePoints = [
         // service-a: high CPU values (should trigger alert)