mgr/cephadm: Add AlertmanagerService

Signed-off-by: Sebastian Wagner <sebastian.wagner@suse.com>
(cherry picked from commit d74f0e9949f72fd8c7793b4435d773d85b910f91)

diff --git a/src/pybind/mgr/cephadm/module.py b/src/pybind/mgr/cephadm/module.py
index 3e89200d3b4705a8d769c220620d9c9234c6312d..5c430476cd4c508eecd14f38bfc32cb84c996fc9 100644 (file)
--- a/src/pybind/mgr/cephadm/module.py
+++ b/src/pybind/mgr/cephadm/module.py
@@ -40,7 +40,7 @@ from .services.cephadmservice import MonService, MgrService, MdsService, RgwServ
     RbdMirrorService
 from .services.nfs import NFSService
 from .services.osd import RemoveUtil, OSDRemoval, OSDService
-from .services.monitoring import GrafanaService
+from .services.monitoring import GrafanaService, AlertmanagerService
 from .inventory import Inventory, SpecStore, HostCache
 
 try:
@@ -429,6 +429,7 @@ class CephadmOrchestrator(orchestrator.Orchestrator, MgrModule):
         self.rgw_service = RgwService(self)
         self.rbd_mirror_service = RbdMirrorService(self)
         self.grafana_service = GrafanaService(self)
+        self.alertmanager_service = AlertmanagerService(self)
 
     def shutdown(self):
         self.log.debug('shutdown')
@@ -1838,7 +1839,7 @@ class CephadmOrchestrator(orchestrator.Orchestrator, MgrModule):
                     self.nfs_service._generate_nfs_config(daemon_type, daemon_id, host)
             extra_args.extend(['--config-json', '-'])
         elif daemon_type == 'alertmanager':
-            cephadm_config, deps = self._generate_alertmanager_config()
+            cephadm_config, deps = self.alertmanager_service.generate_config()
             extra_args.extend(['--config-json', '-'])
         else:
             # Ceph.daemons (mon, mgr, mds, osd, etc)
@@ -1939,7 +1940,7 @@ class CephadmOrchestrator(orchestrator.Orchestrator, MgrModule):
             'rbd-mirror': self.rbd_mirror_service.create,
             'nfs': self.nfs_service.create,
             'grafana': self.grafana_service.create,
-            'alertmanager': self._create_alertmanager,
+            'alertmanager': self.alertmanager_service.create,
             'prometheus': self._create_prometheus,
             'node-exporter': self._create_node_exporter,
             'crash': self._create_crash,
@@ -2422,68 +2423,6 @@ scrape_configs:
         # type: () -> str
         return self.get('mgr_map').get('services', {}).get('dashboard', '')
 
-    def _generate_alertmanager_config(self):
-        # type: () -> Tuple[Dict[str, Any], List[str]]
-        deps = [] # type: List[str]
-
-        # dashboard(s)
-        dashboard_urls = []
-        mgr_map = self.get('mgr_map')
-        port = None
-        proto = None  # http: or https:
-        url = mgr_map.get('services', {}).get('dashboard', None)
-        if url:
-            dashboard_urls.append(url)
-            proto = url.split('/')[0]
-            port = url.split('/')[2].split(':')[1]
-        # scan all mgrs to generate deps and to get standbys too.
-        # assume that they are all on the same port as the active mgr.
-        for dd in self.cache.get_daemons_by_service('mgr'):
-            # we consider mgr a dep even if the dashboard is disabled
-            # in order to be consistent with _calc_daemon_deps().
-            deps.append(dd.name())
-            if not port:
-                continue
-            if dd.daemon_id == self.get_mgr_id():
-                continue
-            addr = self.inventory.get_addr(dd.hostname)
-            dashboard_urls.append('%s//%s:%s/' % (proto, addr.split(':')[0],
-                                                 port))
-
-        yml = """# generated by cephadm
-# See https://prometheus.io/docs/alerting/configuration/ for documentation.
-
-global:
-  resolve_timeout: 5m
-
-route:
-  group_by: ['alertname']
-  group_wait: 10s
-  group_interval: 10s
-  repeat_interval: 1h
-  receiver: 'ceph-dashboard'
-receivers:
-- name: 'ceph-dashboard'
-  webhook_configs:
-{urls}
-""".format(
-    urls='\n'.join(
-        ["  - url: '{}api/prometheus_receiver'".format(u)
-         for u in dashboard_urls]
-    ))
-        peers = []
-        port = '9094'
-        for dd in self.cache.get_daemons_by_service('alertmanager'):
-            deps.append(dd.name())
-            addr = self.inventory.get_addr(dd.hostname)
-            peers.append(addr.split(':')[0] + ':' + port)
-        return {
-            "files": {
-                "alertmanager.yml": yml
-            },
-            "peers": peers
-        }, sorted(deps)
-
     def add_prometheus(self, spec):
         return self._add_daemon('prometheus', spec, self._create_prometheus)
 
@@ -2534,16 +2473,12 @@ receivers:
 
     def add_alertmanager(self, spec):
         # type: (ServiceSpec) -> AsyncCompletion
-        return self._add_daemon('alertmanager', spec, self._create_alertmanager)
+        return self._add_daemon('alertmanager', spec, self.alertmanager_service.create)
 
     @trivial_completion
     def apply_alertmanager(self, spec: ServiceSpec):
         return self._apply(spec)
 
-    def _create_alertmanager(self, daemon_id, host):
-        return self._create_daemon('alertmanager', daemon_id, host)
-
-
     def _get_container_image_id(self, image_name):
         # pick a random host...
         host = None

diff --git a/src/pybind/mgr/cephadm/services/monitoring.py b/src/pybind/mgr/cephadm/services/monitoring.py
index f7d2fda4d7a0afbfb8f3877de4cad9fe5e76877b..f05ff4fe005b05a22f3bb53eacabd7d6d5b0641a 100644 (file)
--- a/src/pybind/mgr/cephadm/services/monitoring.py
+++ b/src/pybind/mgr/cephadm/services/monitoring.py
@@ -102,3 +102,70 @@ datasources:
             }
         }
         return config_file, sorted(deps)
+
+
+class AlertmanagerService(CephadmService):
+    def create(self, daemon_id, host) -> str:
+        return self.mgr._create_daemon('alertmanager', daemon_id, host)
+
+    def generate_config(self):
+        # type: () -> Tuple[Dict[str, Any], List[str]]
+        deps = [] # type: List[str]
+
+        # dashboard(s)
+        dashboard_urls = []
+        mgr_map = self.mgr.get('mgr_map')
+        port = None
+        proto = None  # http: or https:
+        url = mgr_map.get('services', {}).get('dashboard', None)
+        if url:
+            dashboard_urls.append(url)
+            proto = url.split('/')[0]
+            port = url.split('/')[2].split(':')[1]
+        # scan all mgrs to generate deps and to get standbys too.
+        # assume that they are all on the same port as the active mgr.
+        for dd in self.mgr.cache.get_daemons_by_service('mgr'):
+            # we consider mgr a dep even if the dashboard is disabled
+            # in order to be consistent with _calc_daemon_deps().
+            deps.append(dd.name())
+            if not port:
+                continue
+            if dd.daemon_id == self.mgr.get_mgr_id():
+                continue
+            addr = self.mgr.inventory.get_addr(dd.hostname)
+            dashboard_urls.append('%s//%s:%s/' % (proto, addr.split(':')[0],
+                                                 port))
+
+        yml = """# generated by cephadm
+# See https://prometheus.io/docs/alerting/configuration/ for documentation.
+
+global:
+  resolve_timeout: 5m
+
+route:
+  group_by: ['alertname']
+  group_wait: 10s
+  group_interval: 10s
+  repeat_interval: 1h
+  receiver: 'ceph-dashboard'
+receivers:
+- name: 'ceph-dashboard'
+  webhook_configs:
+{urls}
+""".format(
+    urls='\n'.join(
+        ["  - url: '{}api/prometheus_receiver'".format(u)
+         for u in dashboard_urls]
+    ))
+        peers = []
+        port = '9094'
+        for dd in self.mgr.cache.get_daemons_by_service('alertmanager'):
+            deps.append(dd.name())
+            addr = self.mgr.inventory.get_addr(dd.hostname)
+            peers.append(addr.split(':')[0] + ':' + port)
+        return {
+            "files": {
+                "alertmanager.yml": yml
+            },
+            "peers": peers
+        }, sorted(deps)