We are excited to announce support for pausing and resuming Postgres database clusters in our fully managed cloud service, EDB BigAnimal. This new functionality allows organizations to save on compute costs by pausing their clusters as necessary without losing their data or configurations.
Enterprises want their databases to run reliably and be highly available, but not all database clusters need to be running all the time, such as dev/test clusters. In these cases, enterprises want to save costs and only pay for the compute they use.
With BigAnimal, enterprises get even more control and flexibility over their database operations.
Increase flexibility and control
EDB BigAnimal now allows you to hit “pause” on any of your clusters at any time to save on compute costs. It’s easy to start your cluster again by simply clicking a button or running a command to resume operations. You can do this without any potential data loss, and you won’t need to reattach storage volumes or restore from a backup.
In some cases, you may still want to be able to read from a paused cluster. Regardless of the reason, BigAnimal enables you to pause a cluster but continue reading from its faraway replicas. While the cluster is paused, you won’t have to pay for the source cluster’s compute.
Instead of having to reconfigure your apps to run on a new service that separates storage and compute to minimize your costs, you can save money by simply pausing unused clusters.]
Save on compute
Pausing a high availability Postgres cluster can save you money on compute costs in several ways. The capability is available to all BigAnimal customers, and the paused clusters themselves do not incur database or infrastructure costs for their virtual machine, load balancing or networking. This capability supports:
- Reduced CPU and memory usage: When the cluster is paused, the nodes in the cluster are not actively processing any queries, using less CPU and memory resources.
- Reduced disk I/O: Paused PostgreSQL clusters also generate less disk I/O because they are not writing any new data to disk.
- Reduced network traffic: Paused clusters generate less network traffic because they are not communicating with any other nodes in the cluster.
You will still pay infrastructure costs for disk and object storage for paused clusters, because you are still storing data and preserving cluster configurations. The compute cost depends on several factors, such as the size of the cluster, the type of workloads it is running and the cloud provider you are using. By pausing seldom-used clusters, you can save on your compute costs, and by scheduling pauses, you can ensure that you’re in control of your clusters and compute costs. This increases your enterprise’s overall control and flexibility over your database operations, so you can better manage your costs and adopt a FinOps approach.
Configuration and data stores will remain static and unchanged while paused, allowing you to pause your costs but not security.
BigAnimal is the only distributed, fully managed, Oracle-compatible Postgres database. Get started within minutes for free with $300 in cloud credits.