Teach json how to serialize decimals

sample-config/pgmon-metrics.yml

@@ -8,14 +8,21 @@ metrics:
       0: >
         SELECT datname AS dbname
         FROM pg_database
+
+  # Note: If the user lacks sufficient privileges, these fields will be NULL.
+  #       The WHERE clause is intended to prevent Zabbix from discovering a
+  #       connection it cannot monitor.  Ideally this would generate an error
+  #       instead.
   discover_rep:
     type: set
     query:
       0: >
-        SELECT client_addr || '_' || regexp_replace(application_name, '[ ,]', '_', 'g') AS repid,
+        SELECT host(client_addr) || '_' || regexp_replace(application_name, '[ ,]', '_', 'g') AS repid,
                client_addr,
                state
         FROM pg_stat_replication
+        WHERE state IS NOT NULL
+
   discover_slots:
     type: set
     query:
@@ -36,6 +43,7 @@ metrics:
                active
         FROM pg_replication_slots
 
+
   ##
   # cluster-wide metrics
   ##
@@ -85,7 +93,6 @@ metrics:
         FROM pg_stat_bgwriter bg
         CROSS JOIN pg_stat_checkpointer cp
 
-
   io_per_backend:
     type: set
     query:
@@ -108,6 +115,7 @@ metrics:
         FROM pg_stat_io
         GROUP BY backend_type
 
+
   ##
   # Per-database metrics
   ##
@@ -210,14 +218,20 @@ metrics:
 #        FROM pg_class c
 #        CROSS JOIN LATERAL pg_sequence_parameters(c.oid) AS s
 #        WHERE c.relkind = 'S'
-      100000: >
-        SELECT COALESCE(MAX(pg_sequence_last_value(c.oid)::float / sp.maximum_value), 0) AS max_usage
-        FROM pg_class c
-        CROSS JOIN LATERAL pg_sequence_parameters(c.oid) AS sp
-        WHERE c.relkind = 'S'
+      100000: SELECT COALESCE(MAX(last_value::float / max_value), 0) AS max_usage FROM pg_sequences;
     test_args:
       dbname: postgres
 
+  sequence_visibility:
+    type: row
+    query:
+      100000: >
+        SELECT COUNT(*) FILTER (WHERE has_sequence_privilege(c.oid, 'SELECT,USAGE'::text)) AS readable_sequences,
+               COUNT(*) AS total_sequences
+        FROM pg_class AS c
+        WHERE relkind = 'S';
+
+
   ##
   # Per-replication metrics
   ##
@@ -237,7 +251,7 @@ metrics:
                NULL AS replay_lag,
                sync_state
         FROM pg_stat_replication
-        WHERE client_addr || '_' || regexp_replace(application_name, '[ ,]', '_', 'g') = %(repid)s
+        WHERE host(client_addr) || '_' || regexp_replace(application_name, '[ ,]', '_', 'g') = %(repid)s
       100000: >
         SELECT pid, usename,
                EXTRACT(EPOCH FROM backend_start)::integer AS backend_start,
@@ -250,7 +264,8 @@ metrics:
                COALESCE(EXTRACT(EPOCH FROM flush_lag), 0)::integer AS flush_lag,
                COALESCE(EXTRACT(EPOCH FROM replay_lag), 0)::integer AS replay_lag,
                sync_state
-        FROM pg_stat_replication WHERE client_addr || '_' || regexp_replace(application_name, '[ ,]', '_', 'g') = %(repid)s
+        FROM pg_stat_replication
+        WHERE host(client_addr) || '_' || regexp_replace(application_name, '[ ,]', '_', 'g') = %(repid)s
     test_args:
       repid: 127.0.0.1_test_rep
 
@@ -264,24 +279,25 @@ metrics:
       90400: >
         SELECT NULL as active_pid,
                xmin,
-               pg_xlog_location_diff(pg_current_xlog_location(), restart_lsn) AS restart_bytes,
+               pg_xlog_location_diff(pg_current_xlog_location(), restart_lsn)::bigint AS restart_bytes,
                NULL AS confirmed_flush_bytes
         FROM pg_replication_slots WHERE slot_name = %(slot)s
       90600: >
         SELECT active_pid,
                xmin,
-               pg_xlog_location_diff(pg_current_xlog_location(), restart_lsn) AS restart_bytes,
-               pg_xlog_location_diff(pg_current_xlog_location(), confirmed_flush_lsn) AS confirmed_flush_bytes
+               pg_xlog_location_diff(pg_current_xlog_location(), restart_lsn)::bigint AS restart_bytes,
+               pg_xlog_location_diff(pg_current_xlog_location(), confirmed_flush_lsn)::bigint AS confirmed_flush_bytes
         FROM pg_replication_slots WHERE slot_name = %(slot)s
       100000: >
         SELECT active_pid,
                xmin,
-               pg_wal_lsn_diff(pg_current_wal_lsn(), restart_lsn) AS restart_bytes,
-               pg_wal_lsn_diff(pg_current_wal_lsn(), confirmed_flush_lsn) AS confirmed_flush_bytes
+               pg_wal_lsn_diff(pg_current_wal_lsn(), restart_lsn)::bigint AS restart_bytes,
+               pg_wal_lsn_diff(pg_current_wal_lsn(), confirmed_flush_lsn)::bigint AS confirmed_flush_bytes
         FROM pg_replication_slots WHERE slot_name = %(slot)s
     test_args:
       slot: test_slot
 
+
   ##
   # Debugging
   ##

src/pgmon.py

@@ -27,6 +27,8 @@ from urllib.parse import urlparse, parse_qs
 import requests
 import re
 
+from decimal import Decimal
+
 VERSION = "1.0.3"
 
 # Configuration
@@ -391,6 +393,16 @@ def get_query(metric, version):
     raise MetricVersionError("Missing metric query for PostgreSQL {}".format(version))
 
 
+def json_encode_special(obj):
+    """
+    Encoder function to handle types the standard JSON package doesn't know what
+    to do with
+    """
+    if isinstance(obj, Decimal):
+        return float(obj)
+    raise TypeError(f'Cannot serialize object of {type(obj)}')
+
+
 def run_query_no_retry(pool, return_type, query, args):
     """
     Run the query with no explicit retry code
@@ -408,11 +420,11 @@ def run_query_no_retry(pool, return_type, query, args):
                 elif return_type == "row":
                     if len(res) == 0:
                         return "[]"
-                    return json.dumps(res[0])
+                    return json.dumps(res[0], default=json_encode_special)
                 elif return_type == "column":
                     if len(res) == 0:
                         return "[]"
-                    return json.dumps([list(r.values())[0] for r in res])
+                    return json.dumps([list(r.values())[0] for r in res], default=json_encode_special)
                 elif return_type == "set":
                     return json.dumps(res)
         except:

src/test_pgmon.py

@@ -5,6 +5,8 @@ import tempfile
 
 import logging
 
+from decimal import Decimal
+
 import pgmon
 
 # Silence most logging output
@@ -789,3 +791,17 @@ metrics:
         # Make sure we can pull the RSS file (we assume the 9.6 series won't be getting
         # any more updates)
         self.assertEqual(pgmon.get_latest_version(), 90624)
+
+    def test_json_encode_special(self):
+        # Confirm that we're getting the right type
+        self.assertFalse(isinstance(Decimal('0.5'), float))
+        self.assertTrue(isinstance(pgmon.json_encode_special(Decimal('0.5')), float))
+
+        # Make sure we get sane values
+        self.assertEqual(pgmon.json_encode_special(Decimal('0.5')), 0.5)
+        self.assertEqual(pgmon.json_encode_special(Decimal('12')), 12.0)
+
+        # Make sure we can still fail for other types
+        self.assertRaises(
+            TypeError, pgmon.json_encode_special, object
+        )