[AI-5553] Add ha metrics (DataDog#20763)

* add ha metrics

* Add changelog

* Validate metadata

* Add more entries

Author: sarah-witt

proxmox/changelog.d/20763.added

@@ -0,0 +1 @@
+Add HA metrics.

proxmox/datadog_checks/proxmox/check.py

@@ -54,6 +54,22 @@ def _get_vm_hostname(self, vm_id, vm_name, node):
         hostname = hostname_json.get("data", {}).get("result", {}).get("host-name", vm_name)
         return hostname
 
+    def _collect_ha_metrics(self):
+        ha_response = self.http.get(f"{self.config.proxmox_server}/cluster/ha/status/current")
+        ha_response_json = ha_response.json()
+        ha_statuses = ha_response_json.get('data', [])
+        for ha_status in ha_statuses:
+            if not ha_status.get('type') == 'quorum':
+                continue
+            status = ha_status.get('status')
+            quorate = ha_status.get('quorate')
+            status_value = status == "OK"
+            node = ha_status.get('node')
+            tags = [f'node_status:{status}']
+            self.gauge('ha.quorum', status_value, hostname=node, tags=tags)
+            if isinstance(quorate, int):
+                self.gauge('ha.quorate', quorate, hostname=node, tags=tags)
+
     def _collect_performance_metrics(self):
         metrics_response = self.http.get(f"{self.config.proxmox_server}/cluster/metrics/export")
         metrics_response_json = metrics_response.json()
@@ -173,3 +189,4 @@ def check(self, _):
 
         self._collect_resource_metrics()
         self._collect_performance_metrics()
+        self._collect_ha_metrics()

proxmox/metadata.csv

@@ -15,6 +15,8 @@ proxmox.disk.read,count,,,,,0,proxmox,disk read,,
 proxmox.disk.total,gauge,,,,,0,proxmox,disk total,,
 proxmox.disk.used,gauge,,,,,0,proxmox,disk used,,
 proxmox.disk.write,count,,,,,0,proxmox,disk write,,
+proxmox.ha.quorate,gauge,,,,Whether or not the cluster is in quorate,0,proxmox,ha quorate,,
+proxmox.ha.quorum,gauge,,,,Whether or not the node is in quorum,0,proxmox,ha quorum,,
 proxmox.mem,gauge,,,,,0,proxmox,mem,,
 proxmox.mem.max,gauge,,,,,0,proxmox,max mem,,
 proxmox.mem.total,gauge,,,,,0,proxmox,mem total,,

proxmox/tests/common.py

@@ -46,6 +46,8 @@
     'proxmox.swap.used',
 ]
 
+HA_METIRCS = ['proxmox.ha.quorate', 'proxmox.ha.quorum']
+
 NODE_RESOURCE_METRICS = set(RESOURCE_METRICS) - {
     'proxmox.diskread',
     'proxmox.diskwrite',
@@ -77,4 +79,4 @@
 
 STORAGE_PERF_METRICS = {'proxmox.disk.total', 'proxmox.disk.used'}
 
-ALL_METRICS = BASE_METRICS + RESOURCE_METRICS + PERF_METRICS
+ALL_METRICS = BASE_METRICS + RESOURCE_METRICS + PERF_METRICS + HA_METIRCS

proxmox/tests/fixtures/GET/api2/json/cluster/ha/status/current/response.json

@@ -0,0 +1,23 @@
+{
+    "data": [
+        {
+            "status": "OK",
+            "type": "quorum",
+            "node": "ip-122-82-3-112",
+            "quorate": 1,
+            "id": "quorum"
+        },
+        {
+            "status": "OK",
+            "type": "lrm",
+            "node": "ip-122-82-3-112",
+            "id": "lrm"
+        },
+        {
+            "status": "OK",
+            "type": "quorum",
+            "node": "ip-122-82-3-112",
+            "id": "quorum"
+        }
+    ]
+}

proxmox/tests/test_unit.py

@@ -474,3 +474,11 @@ def test_perf_metrics_error(dd_run_check, caplog, instance):
     caplog.set_level(logging.DEBUG)
     dd_run_check(check)
     assert "Invalid metric entry found; metric name: disk.used, resource id: storage/ip-122-82-3-112" in caplog.text
+
+
+@pytest.mark.usefixtures('mock_http_get')
+def test_ha_metrics(dd_run_check, aggregator, instance):
+    check = ProxmoxCheck('proxmox', {}, [instance])
+    dd_run_check(check)
+    aggregator.assert_metric('proxmox.ha.quorum', hostname='ip-122-82-3-112', tags=['node_status:OK'])
+    aggregator.assert_metric('proxmox.ha.quorate', hostname='ip-122-82-3-112', tags=['node_status:OK'])