Microsoft further slashed the price for access to its Azure Cosmos DB distributed database service following a more dramatic price cut earlier this year.
The Azure Cosmos DB platform is designed to provide user control to scale database throughput and storage across regions. Service level agreements (SLAs) are also supported for latency, throughput, availability, and consistency.
The computing and cloud giant lowered the entry point, and thus related cost, for provisioning container request units (RUs) through the platform. The limit change is for partitioned containers.
The previous entry point of unlimited provisioning at 2,500 RUs per second is now set at 1,000 RUs per second. That lower limit maintains the previous scaling level of 100 RU increments, and also the unlimited maximum throughput level.
Microsoft also raised the provisioning limit to 1 million RUs per second per container before triggering support request.
Rimma Nehme, group product manager for the Azure Cosmos DB product, in a blog post said the move results in a 60 percent cost reduction.
“This enables customers to elastically scale throughput based on the application traffic patterns across different geographical regions to support fluctuating workloads varying both by geography and time,” Nehme explained.
Single partition containers allow for as low as 400 RUs per second, scaling up to a maximum of 10,000 RUs per second.
Microsoft in February rolled out a 75 percent price cut that came from moving the entry point from 10,100 RUs per second to the 2,500 RUs per second level.
Microsoft in September bolstered its Azure Cosmos DB platform by integrating its serverless computing-based Azure Functions product. The enhanced product allows for developers to use event-driven serverless computing at scale across the Azure cloud platform.
Different features and functions make it difficult for direct comparisons of management fees between container providers.
For instance, Microsoft recently adjusted pricing for its managed Kubernetes service to only charge for the virtual machines (VMs) that the Kubernetes clusters are running on and not for the actual cluster management.
Google last month eliminated the cluster management fee for its Google Kubernetes Engine (GKE), regardless of the number of clusters being managed. The company had been charging a $0.15 per hour fee to manage a cluster with more than six nodes. Customers are still charged for each node in a cluster based on its resource size and length of use on a per-second basis with a one minute minimum usage cost.
Amazon Web Services (AWS) through its Elastic Cloud Compute (EC2) model only charges for resources used to create and run an application within a container, but does not charge for its Amazon Elastic Container Service (ECS). AWS more recently launched its Fargate model, which allows customers to run containers without managing servers and clusters.