Google scales Cloud Bigtable NoSQL database

0

Google today announced a series of generally available feature updates to its database service, Cloud Bigtable, designed to help improve scalability and performance.

Google Cloud Bigtable is a managed NoSQL database service capable of handling both analytical and operational workloads. New updates Google is bringing to Bigtable include increased storage capacity, with up to 5TB of storage per node now available, an increase from the previous limit of 2.5TB. Google also offers improved autoscaling capabilities, so a database cluster automatically grows or shrinks as needed. Rounding out the Bigtable update is improved visibility into database workloads to enable better troubleshooting of issues.

The new features announced in Bigtable demonstrate the continued focus on increased automation and extension that are becoming tabletop stakes for modern cloud services.

Adam RonthalAnalyst, Gardener

“The new features announced in Bigtable demonstrate the continued focus on increased automation and extensibility that is becoming the tabletop for modern cloud services,” he said Adam Ronthal, Analyst at Gartner. “They also advance the goal of improved price and performance—which is fast becoming the key metric for evaluating and managing cloud services—and observability, which serves as the foundation for improved financial management and optimization.”

How automatic scaling is transforming Google Cloud Bigtable’s database operations

ONE promise of the cloud has long been the ability to elastically scale resources on demand without requiring new physical infrastructure for end users.

According to Anton Gething, Bigtable product manager at Google, programmatic scaling has always been available in Bigtable. He added that many Google customers have developed their own autoscaling approaches for Bigtable via the programmatic APIs. For example, Spotify has provided an open-source implementation for Cloud Bigtable’s auto-scaling.

“Today‘s Bigtable release introduces a native autoscaling solution,” Gething said.

He added that the native autoscaling monitors the Bigtable servers directly to be very responsive. As demand changes, the size of a Bigtable deployment can also change.

The size of each Bigtable node will also be increased in the new update. Previously, Bigtable had a maximum storage capacity of 2.5 TB per node; that’s now doubled to 5TB.

According to Gething, users don’t need to upgrade their existing deployment to take advantage of the increased storage capacity. He added that Bigtable has separated compute and storage functions, allowing each type of resource to scale independently.

“This memory capacity refresh is intended to provide cost optimization for memory-driven workloads that require more memory without increasing compute power,” Gething said.

Optimizing Google Cloud Bigtable database workloads

Another new feature that has landed in Bigtable is a feature called Cluster group routing.

Gething explained that cluster groups in a replicated Cloud Bigtable instance provide finer-grained control over high-availability deployments and improved workload management. Before the new update, he found that a user of a replicated Bigtable instance could route traffic to one of its Bigtable clusters in a single-cluster routing mode, or to all of its clusters in a multi-cluster routing mode. He said cluster groups now allow customers to route traffic to a subset of their clusters.

Google also added a new CPU Utilization by App Profile metric that provides more insight into the performance of a specific application workload. While Google did give Bigtable users some insight into CPU usage prior to the new update, Gething explained that the new update brings new dimensions of visibility to the data query access methods and to the database tables that were accessed.

“Before those extra dimensions, it could be difficult to troubleshoot,” Gething said. “You would have visibility into cluster CPU usage, but you wouldn’t know which app profile traffic was consuming CPU or which table was being accessed and by what method.”

Share.

Comments are closed.