Google rolled out a new cross-platform API for enabling faster access to the company's Cloud Datastore NoSQL database from outside Google's App Engine.
The new Cloud Datastore API, currently available as a beta release for Google's cloud customers, is designed to improve the performance and reliability of the database for those accessing it from Google's Container Engine, Compute Engine and other platforms, Dan McGrath, Google's cloud platform product manager, announced.
Google also released details of the service-level agreement (SLA) that will take effect when the new API becomes generally available, but the company hasn't said when that would happen. The SLAs for the new API cover both the company's Cloud Datastore multi-region service and its Cloud Datastore regional service.
As with Google's other SLAs, the company will guarantee a monthly uptime of at least 99.95 percent for customers of the Cloud Datastore multi-region service. That works out to roughly 22 minutes of downtime per month or about 4 hours and 23 minutes per year. For customers of the Cloud Datastore regional service, Google will offer uptime greater than or equal to 99.9 percent, which translates to slightly less than 44 minutes of monthly downtime or about 8 hours and 46 minutes per year.
Google will offer varying amounts of financial credit to customers if it fails to deliver on the guaranteed service-level obligations. For instance, multi-region Cloud Datastore customers will receive 10 percent off their monthly bill if service uptime falls below the guaranteed 99.95 percent, but remains above 99 percent. Customers will receive bigger credits for bigger service disruptions, topping off at 50 percent for service uptimes that fall below 95 percent, or a disruption of about one and a half days per month. Similar financial credits will be available for customers of Cloud Datastore regional as well.
Google Cloud Datastore is a NoSQL Web and mobile application database for the company's cloud customers. The product is one of several technologies the company offers as a storage service for enterprises and consumers. Others include Google Cloud Storage for unstructured data, Cloud SQL for structured data, Cloud Bigtable for high-volume data and Google Drive for personal storage purposes.
Google's Cloud Datastore API announcement comes just one week after the company announced a simplified pricing model for the service that it claimed would result in lower costs for most current customers.
The new pricing model is scheduled to go into effect July 1 and will simplify the manner in which database read, write and delete actions are counted for pricing purposes, according to Google. The changes will address some of the disincentives in the current model for using some of the more powerful features in the Cloud Datastore technology, the company noted in announcing the new pricing structure.
In addition, Google introduced a new method for calculating stored data in Datastore so developers can more accurately estimate storage costs. Like the pricing changes, the new method for calculating bytes stored will reduce storage costs for a majority of customers, Google claimed.