Tag FoundationDB process metrics with process class, assigned roles (#19682)

* Tag FoundationDB process metrics with process class, assigned roles

* Test that FoundationDB process metrics are tagged with process class and role

Author: jon-signal

foundationdb/changelog.d/19682.added

@@ -0,0 +1 @@
+Tag FoundationDB process metrics with process class, assigned roles

foundationdb/datadog_checks/foundationdb/check.py

@@ -93,6 +93,19 @@ def report_process(self, process):
             return
         tags = self._tags + ["fdb_process:" + process["address"]]
 
+        if "class_type" in process:
+            tags.append("fdb_process_class:" + process["class_type"])
+
+        if "roles" in process:
+            # First, report role-specific metrics before we add role tags to the list…
+            for role in process["roles"]:
+                self.report_role(role, tags)
+
+            # …then add role tags so we can associate process metrics with roles.
+            for role in process["roles"]:
+                if "role" in role:
+                    tags.append("fdb_role:" + role["role"])
+
         if "cpu" in process:
             self.maybe_gauge("process.cpu.usage_cores", process["cpu"], "usage_cores", tags)
         if "disk" in process:
@@ -122,10 +135,6 @@ def report_process(self, process):
             self.maybe_hz_counter("process.network.megabits_sent", network, "megabits_sent", tags)
             self.maybe_hz_counter("process.network.tls_policy_failures", network, "tls_policy_failures", tags)
 
-        if "roles" in process:
-            for role in process["roles"]:
-                self.report_role(role, tags)
-
     def report_role(self, role, process_tags):
         if "role" not in role:
             return

foundationdb/tests/test_foundationdb.py

@@ -36,6 +36,29 @@ def test_full(aggregator, instance):
             aggregator.assert_metric(metric)
             aggregator.assert_metric_has_tag(metric, 'fdb_test:true')
 
+            if metric.startswith('foundationdb.process.'):
+                aggregator.assert_metric_has_tag_prefix(metric, 'fdb_process_class')
+                aggregator.assert_metric_has_tag_prefix(metric, 'fdb_role')
+
+        aggregator.assert_metric(
+            'foundationdb.process.cpu.usage_cores',
+            tags=[
+                'fdb_test:true',
+                'fdb_process:127.0.0.1:4000',
+                'fdb_process_class:unset',
+                'fdb_role:master',
+                'fdb_role:cluster_controller',
+                'fdb_role:data_distributor',
+                'fdb_role:ratekeeper',
+                'fdb_role:coordinator',
+                'fdb_role:proxy',
+                'fdb_role:log',
+                'fdb_role:storage',
+                'fdb_role:resolver',
+            ],
+            value=0.015280199999999999,
+        )
+
         aggregator.assert_all_metrics_covered()
         aggregator.assert_metrics_using_metadata(get_metadata_metrics())
         aggregator.assert_service_check("foundationdb.can_connect", AgentCheck.OK)