The Azure Cosmos DB for PostgreSQL cluster compute start/stop feature allows you to completely stop compute on all nodes in a Postgres cluster and pick up later where you left off, saving time and costs.
Previously, you would need to run continuously clusters that are needed only from time to time such as dev/test environment. Or you would need to delete a cluster and provision a new one later requiring potentially time-consuming data ingestion and cluster configuration.
The cluster compute start/stop feature allows to keep all cluster data and configuration as is for unlimited period without cost of cluster compute. Then you can start compute on all cluster nodes and get your cluster in a state when compute was stopped.
Learn More about compute start stop in Azure Cosmos DB for PostgreSQL