IBM wipes away tiers to join cloud storage price wars

IBM's decided to play the “our cloud storage is even cheaper than your cloud storage” game, but by a different set of rules.

Big Blue's decided that tiered cloud storage assumes users make decisions based on unusually-potent-and-accurate insights into their long-term data use patterns, and therefore commit themselves to a tier of storage in the hope it's the cheapest place to store that data. But IBM reckons modern analytics and cognitive tools mean you'll probably start to consider archival or near-archival data more often, which it thinks will make you look pretty foolish if you have to pay to haul a heap of it out of AWS Glacier or Google Nearline to do some work.

IBM's therefore cooked up what it calls “Flex” storage. As the draft pricing table below shows, the product offers a flat price for cloud storage regardless of how much you access it. Big Blue's betting that over time you'll end up ahead, even if the per-gigabyte-per-month price is higher than that offered by rivals. The company also thinks you'll appreciate just tossing data into the cloud, rather than having to ponder just what tier and just what region to choose.

Draft pricing for IBM Flex cloud storage. Big Blue may change these numbers before launching in early Q2 2017

Big Blue's been busy on the storage front with two other deals. One will see NetApp's AltaVault product gain the ability to send backups to IBM Cloud Object Storage. That data could lodge in the new Cold Vault IBM's added to its cloudy Object Storage service. Cold Vault is akin to AWS Glacier.

IBM's third storage move is a deal with Veritas that will see NetBackup 8.0 join the Bluemix Catalog of cloudy software. Veritas' flagship will also be able to use IBM's cloudy object storage as a backup target, an addition touted as helping to migrate data from on-premises storage to the cloud.

Also announced at IBM's aptly named “InterConnect” conference is a new deal with Red Hat, which has signed up IBM as a Certified Cloud and Service Provider. The practical effect of that relationship is that you'll be able to run Red Hat's OpenStack Platform on IBM's forthcoming Private Cloud. Red Hat's cut of Ceph will also run just fine on Big Blue's boxen. The deal also means that Red Hat customers can use their licences in IBM's cloud to build hybrid clouds.

IBM's also decided that because many organisations are already using multiple cloud or SaaS services, it might as well try to pipe data out of them all and into the maw of its analytics services. The idea here is that individual clouds probably produce decent insights, but that combining data from multiple sources will probably do better. IBM's therefore offering both plumbing and automation to get this stuff happening, some of it helped along by Watson to make sure the right data flows to the right places and that infrastructure can understand when these flows might send some stress its way. ®