Google recently announced various improvements to Cloud Spanner, its distributed, decoupled relational database service with a "50% increase in throughput and 2.5 times the storage per node than before" without a price change.
The company will increase the Cloud Spanner's read throughput, meaning that each Spanner node, i.e., collections of compute resources, namely CPUs, RAM, and storage, can accommodate ten terabytes of storage compared to 4TB. These updates will be rolled out in the coming months, while in some select regional and multi-region instance configurations, it is available now.
The company claims "Spanner's high throughput, virtually unlimited scale, single-digit millisecond latency, five 9s availability SLA, and strong external-consistency semantics are now available at half the cost of Amazon DynamoDB for most workloads." The cost comparison with Dynamo DB is based on numbers from the AWS Prime Day blog post, with authors of a Google blog post stating:
With these changes, Spanner now offers up to 2x better read throughput per dollar compared to Amazon DynamoDB for similar workloads.
The claim and comparison led to various comments online. With Corey Quinn, a chief cloud economist at DuckbillGroup, tweeting:
When you post like this, you just paint a massive target on your back.
Is Google really boasting here that their global transaction volume for Cloud Spanner is "over 20x" DynamoDB's volume for a single (admittedly large, internal) customer?
How the hell did this get approved?
Furthermore, Dax Raad, a founder of Bumi, tweeted:
stuff like this is why i think google cloud is a joke
they took a stat amazon shared about how much traffic one of their own products using publicly available dynamodb does
and compared it to total usage of 3 products using a google-internal version of Spanner
Confirmed by a respondent on a Hacker News thread who wrote:
What would have perhaps been a more fair comparison is to share the peak load that Google services running Cloud Spanner, and not the sum of all Spanner services across all of GCP and all of Google (Spanner on non-GCP infra).
And:
What is confusing, however, is how in this article "Cloud Spanner" is consistently used... except for when talking about Gmail, Ads and Photos, where it's stated that "Spanner" is used by these products, not "Cloud Spanner!". Like if they were not using the Cloud Spanner infra, but their own. It would help to know what is the case, and what the load of Cloud Spanner is: and not Spanner running on internal Google infra that is not GCP.
Lastly, users of Spanner will, according to the company, have to do anything to take advantage of these improvements and will continue to be billed at their current rate. Details of the Cloud Spanner pricing are available on the pricing page.