mgr/prometheus: Fix pg_* counts

Currently, the pg_* counts are not computed properly. We split the
current state by '+' sign but do not add the pg count to the already
found pg count. Instead, we overwrite any existing pg count with the new
count. This patch fixes it by adding all the pg counts together for all
the states.

It also introduces a new pg_total metric for pg_total that shows the
total count of PGs.

Signed-off-by: Boris Ranto <branto@redhat.com>

diff --git a/src/pybind/mgr/prometheus/module.py b/src/pybind/mgr/prometheus/module.py
index 945a6b5a3364773b74675ca3d36792f8221ddcd4..3770a7202dad0452355d317930c2fc95a696eb43 100644 (file)
--- a/src/pybind/mgr/prometheus/module.py
+++ b/src/pybind/mgr/prometheus/module.py
@@ -195,6 +195,13 @@ class Module(MgrModule):
             'POOL Metadata',
             POOL_METADATA
         )
+
+        metrics['pg_total'] = Metric(
+            'gauge',
+            'pg_total',
+            'PG Total Count'
+        )
+
         for state in OSD_STATUS:
             path = 'osd_{}'.format(state)
             self.log.debug("init: creating {}".format(path))
@@ -269,16 +276,25 @@ class Module(MgrModule):
 
     def get_pg_status(self):
         # TODO add per pool status?
-        pg_s = self.get('pg_summary')['all']
-        reported_pg_s = [(s,v) for key, v in pg_s.items() for s in
-                         key.split('+')]
-        for state, value in reported_pg_s:
+        pg_status = self.get('pg_status')
+
+        # Set total count of PGs, first
+        self.metrics['pg_total'].set(
+            pg_status['num_pgs'],
+        )
+
+        reported_states = {}
+        for pg in pg_status['pgs_by_state']:
+            for state in pg['state_name'].split('+'):
+                reported_states[state] =  reported_states.get(state, 0) + pg['count']
+
+        for state in reported_states:
             path = 'pg_{}'.format(state)
             try:
-                self.metrics[path].set(value)
+                self.metrics[path].set(reported_states[state])
             except KeyError:
                 self.log.warn("skipping pg in unknown state {}".format(state))
-        reported_states = [s[0] for s in reported_pg_s]
+
         for state in PG_STATES:
             path = 'pg_{}'.format(state)
             if state not in reported_states: