Added TOC entries and fixed H2 headings in concepts.

Author: niklarin

articles/cosmos-db/postgresql/TOC.yml

@@ -105,6 +105,8 @@
           href: concepts-maintenance.md
         - name: Burstable compute
           href: concepts-burstable-compute.md
+        - name: Start and stop cluster compute
+          href: concepts-compute-start-stop.md
         - name: Upgrade
           href: concepts-upgrade.md
     - name: Business continuity
@@ -145,6 +147,8 @@
           href: howto-create-users.md
         - name: Schedule maintenance
           href: howto-maintenance.md
+        - name: Start and stop cluster compute
+          href: how-to-start-stop-cluster.md
         - name: Restart cluster
           href: howto-restart.md
         - name: Upgrade cluster

articles/cosmos-db/postgresql/concepts-compute-start-stop.md

@@ -17,13 +17,13 @@ Azure Cosmos DB for PostgreSQL allows you to stop compute on all nodes in a clus
 > [!NOTE]
 > Billing for provisioned storage on all cluster nodes continues when cluster's compute is stopped.
 
+## Managing compute state on cluster nodes
+
 You can stop compute on a cluster for as long as you need.
 
 You can perform management operations such as compute or storage scaling, adding a worker node, or updating networking settings only on clusters with started compute. 
 
-If cluster has [high availability (HA)](./concepts-high-availability.md) enabled, compute start and stop operations would start and stop compute on all primary and standby nodes in the cluster.
-
-You can start and stop compute on the primary cluster and any of its [read replicas](./concepts-read-replicas.md) independently.
+If cluster has [high availability (HA)](./concepts-high-availability.md) enabled, compute start and stop operations would start and stop compute on all primary and standby nodes in the cluster. You can start and stop compute on the primary cluster and any of its [read replicas](./concepts-read-replicas.md) independently.
 
 ## Next steps