As more mainstream enterprise applications move to the cloud, Google Cloud today announced two new services to simplify data management and protect data in containers, as well as extensions to its dual-region cloud storage offering.

Users are no longer just running stateless applications in containers. Instead, more workloads are becoming stateful. To help accelerate this, Google Cloud introduced the new native backup service for Google Kubernetes Engine (GKE).

“When people are modernizing their applications, they are coming to GKE,” said Guru Pangal, GM of Storage at Google Cloud. “They are asking us to protect their data, especially stateful workloads.” Google plans to continue investing in GKE, as well as enterprise-grade data protection and availability, he added.

The move comes as GKE becomes one of Google Cloud’s “highest growth areas,” Pangal said. “We have been leading in this space. And we are seeing very strong adoption. It's a rocketship for us.” 

Google Cloud also released Filestore Enterprise, a new addition to the Filestore family. This file management service caters to traditional tier-one enterprise applications like SAP.

Pangal explained that the new cloud service offers a similar-but-faster experience compared to on-premise file storage and functionality, especially for lift and shift applications. 

In addition to high-performance reads and writes, Filestore Enterprise also provides “cross-zone sync replication” within a region. If there is a failure on one zone, Filestore will synchronously replicate the data to the other zone without any operational intervention while continuing to transparently serve data to the application.

One of Google Cloud Storage’s most important features is dual- and multi-region buckets, which enable a “continent-sized storage system,” according to Pangal. “​It's like one bucket that spans across regions,” he said.

To enable developers to treat this as a single bucket, Google Cloud uses technologies such as Colossus, a globally distributed file system, and Spanner, a provable distributed database. The functionality simplifies the application programming model.

“The capability that we have with dual- and multi-region buckets is really unique in the industry,” said Brian Schwarz, director of product management at Google Cloud. Customers can now custom select regions for their buckets, and they also have the option to add a service level agreement-backed 15-minute recovery point objective.

Schwarz cited media companies as an example. The custom selection enables them to choose regions close to their studios and data centers, which could enhance their ability to utilize the cloud. 

“What we were doing so far is we were picking the regions for our customers,” Pangal added. “Now, we want to give the flexibility for our customers to pick their own region.”

Another way to extend their dual-region buckets is to offer “turbo replication.” 

Pangal explained that the optional function can replicate 100% of users’ data between regions in 15 minutes or less. “The industry standard is to copy 99%, and the rest — the 1% — may reach there at some point, that just doesn't cut it when you are moving enterprise applications,” he added. 

Schwarz said the new features will enable customers to replace painful DIY processes of setting up and managing their storage, networking, and operations. 

“Customers have tried to build some of this stuff themselves, like for backups and protecting their own data, and we really reduce the operational toil,” he said. “But to a larger extent, I really think these are opening up new opportunities and workloads.”