Google unveiled a new beta cloud storage offering named Nearline today that aims squarely at the company's biggest rival in the cloud, Amazon Web Services.
Nearline is a public IaaS cloud service that allows customers to store data at a low cost with fast retrieval (in seconds). It competes most directly with AWS's Glacier "cold" storage service, which promises customers inexpensive storage in exchange for long wait times (up to hours) to access data.
+MORE AT NETWORK WORLD: The myth about how Amazon's cloud started just won't die +
The move shows that Google is serious about the IaaS cloud market and that AWS, while leading the market, will face intense competition from providers like Google and Microsoft.
Google engineered Nearline to be an inexpensive storage platform for data with fast response times. It's not an object-storage system that would provide storage for virtual machines , though (Google has its Cloud Storage service for that). Instead, like AWS's Glacier, Nearline will most likely be a good fit for storing data that is not accessed regularly, such as in disaster recovery and backup scenarios. Here are some of the technical specifications:
- Google Nearline costs $0.01 per GB for data at rest
- Average of 3 second response time for data requests
- Automatic data redundancy across multiple physical locations
- Integrated application program interfaces (APIs) with other Google Cloud Platform Services
What Nearline really means
Google's release of Nearline puts pressure on AWS. As the market-leading cloud platform, AWS has a variety of storage offerings, from its popular S3 storage service to its Elastic Block Storage to accompany virtual machines, to Glacier – the company's cold storage offering.
Nearline will compete almost directly with Glacier. Whereas Glacier advertises response times of "several hours," Nearline will retrieve data in seconds. It's a big move for the industry, but Gartner cloud research director Kyle Hilgendorf says it's not a fatal blow to Glacier and AWS. The company already has a huge lead it is competing from.
"Although Nearline appears to be highly competitive or more appealing than Glacier – I also do not expect to see mass migrations off of Glacier and onto Nearline. For customers that are already GCP customers – this is a great add. For AWS customers, it's probably not differentiating enough to warrant a big migration. For customers of either or neither provider and for those that have a very specific backup, archive, DR or static content hosting, Nearline could tip the scales in Google's favor. But always remember, storage rarely lives on an island in a provider and customers like to wrap many other services around the storage. So the lead that AWS has on most of their other cloud services will continue to naturally encourage those customers to use S3 and Glacier." - Kyle Hilgendorf
A little help from friends
Along with announcing the new service today, a handful of partners also jumped on the Nearline bangwagon. Companies like Veritas/Symantec, Iron Mountain, NetApp and Geminare will provide backup services based on Nearline, allowing customers to use those storage management platforms on the front end and Nearline on the back end. Geminare has a DRaaS (Diaster Recovery as a Service) offering based on Nearline out today. Google also has an interface for users to consume Nearline directly.
Google beefs up
Nearline helps to build out Google Cloud Platform's catalog of cloud services. In the storage arena alone, Google now has four options for customers: a default object storage service named Cloud Storage (compare this to AWS's S3), a SQL database and a NoSQL database, and now Nearline.
Hilgendorf says one advantage of Google's platform compared to AWS is that Google maintains consistent APIs for all storage services. In AWS, the S3 and Glacier APIs are similar, but there are differences.
That means Nearline can be used for some innovative new use cases beyond just the natural examples of disaster recovery and backup. Hilgendorf says Nearline could support a "policy-based storage" use case, where data that is infrequently accessed would be stored in Nearline. If that data suddenly begins to be accessed more frequently, then the data could automatically be transferred into Cloud Storage, using the same APIs. That is technically feasible with AWS, but transferring from Glacier to S3 could take hours and requires slightly different API calls.