Bring metric class names in line with those in the code (neo4j#1361)

modules/ROOT/pages/monitoring/metrics/reference.adoc

@@ -354,16 +354,16 @@ The total number of queries executed remotely to a member of a different cluster
 == Metrics specific to clustering
 
 [[catchup-metrics]]
-.Catchup metrics
+.CatchUp metrics
 
 [options="header",cols="<3m,<4"]
 |===
 |Name |Description
 |<prefix>.cluster.catchup.tx_pull_requests_received|TX pull requests received from other cluster members. (counter)
 |===
 
-[[discovery-db-primary-metrics]]
-.Discovery database primary metrics
+[[discovery-metrics]]
+.Discovery metrics
 
 [options="header",cols="<3m,<4"]
 |===
@@ -376,8 +376,8 @@ The total number of queries executed remotely to a member of a different cluster
 |<prefix>.cluster.discovery.restart.failed_count|Discovery restart failed count. (gauge)
 |===
 
-[[raft-db-primary-metrics]]
-.Raft database primary metrics
+[[raft-metrics]]
+.Raft metrics
 
 [options="header",cols="<3m,<4"]
 |===
@@ -457,7 +457,7 @@ You can limit this memory using JVM settings, but there is never any good reason
 |===
 
 [[jvm-heap-metrics]]
-.JVM heap metrics
+.JVM Heap metrics
 
 [options="header",cols="<3m,<4"]
 |===