Have an extra petabyte of data sitting around? Today the company that has already amassed a huge chunk of the world’s data has made it easier to sock away a whole lot more.
This morning Google announced what it’s calling Google Cloud Storage Nearline, a cloud-based version of an old-fashioned way to store massive amounts of data too important to delete but not important enough to keep close.
The service is designed for companies with very large storage requirements — on the order of hundreds of terabytes. Traditionally, businesses that want to archive data at that scale have turned to offline or “cold” storage, meaning that the data isn’t immediately available, usually because it’s stored on tape drives or other media and filed away in a vault somewhere. That saves costs by reducing both electrical use and the number of servers that employees have to maintain. But it means that anytime someone actually wants to retrieve any of the data stored on those drives, someone has to go and find the data and bring it back online.
For Google, however, the main point of the service is that it might help woo more customers from Amazon.
Nearline is meant to sit between these two options. It’s faster than offline storage — Google claims data stored in the system can be retrieved in seconds. But at only one cent per gigabyte per month, not counting retrieval costs, it’s cheaper than most comparable services. Instead, it’s closer to what you’d pay for consumer storage services like Dropbox or Google Drive, which are designed to handle much smaller amounts of data. Google’s regular Cloud Storage service, like Amazon’s similar offering, S3, cost at least $.026 per gig per month.
Nearline’s main use will probably be for creating off-site backups of critical information, and Google has announced partnerships with more traditional storage companies help customers pipe data into Nearline, including Veritas, NetApp and Iron Mountain. But by making cold storage easier and faster to access, Nearline could also open up new possibilities in the world of mass storage.
Better Than A Jukebox
Nearline won’t work well for information that needs to be displayed in near real-time, but for things that users can wait a moment for, it could be perfect. Facebook, for example, built its own method for quickly retrieving older, seldom accessed user photos from cold storage by storing them on BlueRay discs managed by what amounts to a gigantic robotic jukebox system. Other companies will now be able to take advantage of similarly rapid access to archival information for their own applications, without the need for complex robotics.
For Google, however, the main point of the service is that it might help woo more customers from Amazon and other cloud services, says Henry Baltazar, a storage industry analyst at Forrester.
Amazon offers its own cold storage service Glacier. Glacier also costs 1 cent per gigabyte, but it takes three to five hours to retrieve data, and there are steep penalties for trying to retrieve too much data at once. Google hasn’t yet revealed the costs for retrieving data, but the faster speeds alone will bring some much needed competition to the market. “This is the first time we’ve seen a real competitor to Glacier” Baltazar says. “It’s coming in at the same price, but eliminating the bottlenecks.”
Most importantly, though, it could be a way for Google to lure in new customers before selling other cloud services, such as analytics, to them. Says Baltazar: “They’re willing to make it a loss leader so they can make money on the real things that they want to sell.”
No comments:
Post a Comment