Posts Tagged ‘IBM Cloud Object Storage’

IBM Changes the Economics of Cloud Storage

March 31, 2017

Storage tiering used to be simple: active data went to your best high performance storage, inactive data went to low cost archival storage, and cloud storage filled in for one or whatever else was needed. Unfortunately, today’s emphasis on continuous data analytics, near real-time predictive analytics, and now cognitive has complicated this picture and the corresponding economics of storage.

In response, last week IBM unveiled new additions to the IBM Cloud Object Storage family. The company is offering clients new choices for archival data and a new pricing model to more easily apply intelligence to unpredictable data patterns using analytics and cognitive tools.

Analytics drive new IBM cloud storage pricing

By now, line of business (LOB) managers, having been exhorted to leverage big data and analytics for years, are listening. More recently, the analytics drumbeat has expanded to include not just big data but sexy IoT, predictive analytics, machine learning, and finally cognitive science. The idea of keeping data around for a few months and parking it in a long term archive to never be looked at again until it is finally deleted permanently just isn’t happening as it was supposed to (if it ever did). The failure to permanently remove expired data can become costly from a storage standpoint as well as risky from an e-discovery standpoint.

IBM puts it this way: Businesses typically have to manage across three types of data workloads: “hot” for data that’s frequently accessed and used; “cool” for data that’s infrequently accessed and used; and “cold” for archival data. Cold storage is often defined as cheaper but slower. For example, if a business uses cold storage, it typically has to wait to retrieve and access that data, limiting the ability to rapidly derive analytical or cognitive insights. As a result, there is a tendency to store data in more expensive hot storage.

IBM’s new cloud storage offering, IBM Cloud Object Storage Flex (Flex), uses a “pay as you use” model of storage tiers potentially lowering the price by 53 percent compared to AWS S3 IA1 and 75 percent compared to Azure GRS Cool Tier.2 (See footnotes at the bottom of the IBM press release linked to above. However IBM is not publishing the actual Flex storage prices.) Flex, IBM’s new cloud storage service, promises simplified pricing for clients whose data usage patterns are difficult to predict. Flex promises organizations will benefit from the cost savings of cold storage for rarely accessed data, while maintaining high accessibility to all data.

Of course, you could just lower the cost of storage by permanently removing unneeded data.  Simply insist that the data owners specify an expiration date when you set up the storage initially. When the date arrives in 5, 10, 15 years automatically delete the data. At least that’s how I was taught eons ago. Of course storage costs orders of magnitude less now although storage volumes are orders of magnitude greater and near real-time analytics weren’t in the picture.

Without the actual rates for the different storage tiers you cannot determine how much Storage Flex may save you.  What it will do, however, is make it more convenient to perform analytics on archived data you might otherwise not bother with.  Expect this issue to come up increasingly as IoT ramps up and you are handling more data that doesn’t need hot storage beyond the first few minutes of its arrival.

Finally, the IBM Cloud Object Storage Cold Vault (Cold Vault) service gives clients access to cold storage data on the IBM Cloud and is intended to lead the category for cold data recovery times among its major competitors. Cold Vault joins its existing Standard and Vault tiers to complete a range of IBM cloud storage tiers that are available with expanded expertise and methods via Bluemix and through the IBM Bluemix Garages.

DancingDinosaur is Alan Radding, a veteran information technology analyst, writer, and ghost-writer. Please follow DancingDinosaur on Twitter, @mainframeblog. See more of his IT writing at and here.


IBM Unveils Enhanced Repackaged Spectrum Storage

February 9, 2017

IBM appears to be gaining traction with its growing Spectrum SDS family of storage products re-introduced this week. According to the company, 87 of the Fortune Global 100 use IBM Spectrum Storage. That breaks down to all 10 of the top 10 telecommunications companies and all 20 of the top 20 banks. In addition, 18 of the top 20 energy companies, 9 of the top 10 global healthcare companies, and 8 of the top 10 automobile manufacturers adopted Spectrum storage. In addition, IBM notes, 80 organizations pick IBM Spectrum storage every week.

Of course, that hasn’t been enough to turn incessant red ink into black. According to IBM’s 2016 year-end financials, systems (systems hardware—including storage—and operating systems software), posted revenues of $2.5 billion, down 12.5 percent. You can see DancingDinosaur’s report on the latest IBM financials here. Although IBM called out the z for gross profit margins improvements driven by z Systems performance there was nary a word about storage. Will follow upcoming quarterly reports to see if this increased traction translates into actual positive revenue. Stay tuned.

Over the shoulder shot of a group of business colleagues in a meeting around a conference table

IBM introduces Spectrum Computing, 6/16

The announcements this week included:

  • IBM Spectrum Storage Suite
  • IBM Spectrum Virtualize
  • IBM Spectrum Control
  • IBM Spectrum Accelerate
  • IBM Cloud Object Storage

IBM Spectrum Storage isn’t completely new. DancingDinosaur first covered the Spectrum storage introduction in mid-February, 2015. Actually IBM began offering SDS products in 2014 and gained some kudos for it from IDC. The latest announcement really amounts to a repackaging of the products as the IBM Spectrum Storage Suite along with a variety of enhancements, some of which are quite interesting.

For example, IBM Cloud Object Storage software allows new use cases and enables a standalone object store managed by IBM Spectrum Control. It also adds a new storage tier behind IBM Spectrum Scale and a primary pool target behind IBM Spectrum Protect in the form of a cloud container. IBM also continues its innovative licensing arrangement by which you pay for your storage capacity and then can allocate and re-allocate that capacity freely.

Spectrum Cloud Object storage also introduces unified NFS/Object access. This allows companies to store data in a file system structure on object storage using NFS access capability and access data stored as files via either a file or object interface. It has been optimized for scalability and file-to-object migration as well as being able to scale to millions of users and buckets/containers. Finally, it now supports IPv6 management of devices and all nodes in configuration.

IBM Spectrum Virtualize Software also is interesting. For example, it now supports Supermicro SuperServer 2028U-TRTP+ in addition to existing support for Lenovo System x3650 M5. IBM envisions service providers and enterprises deploying Supermicro servers to build new services based on IBM Spectrum Virtualize software to deliver virtualized storage services at a lower price point. Take note: both of these are 2u x86 boxes. They can also offer disaster recovery as a service for clients with SVC.

Finally, IBM has enhanced Spectrum Control in V5.2.13. Among the new capabilities: improved storage insights through new cloud-based storage analytics for Dell EMC VNX, VNXe, and VMAX. This should enable users to improve application performance and reduce storage costs. It also will extend capacity planning views include external storage for IBM Spectrum Scale’s transparent cloud tiering. For on-premises software the latest Spectrum Control offers new support for Dell EMC VNXe file storage.

Overall, the new Spectrum Control should simplify the life of storage managers. “IBM Spectrum Control gives me one pane of glass to manage spinning disk, file system clusters, and object storage, “said Bob Oesterlin, Sr. Principal Storage Engineer, Nuance, as reported by IBM. The ability to span IBM storage as well as that of other vendors should prove a winner.

Combined with other capabilities, such as Spectrum Accelerate V11.5.4’s data-at-rest encryption, ability to flexibly encrypt existing hot data in minutes without disruption, and support for standard key management tools (IBM Security Key Lifecycle Manager and SafeNet KeySecure) will add to the appeal of the enhanced IBM Spectrum Storage Suite. Will it be enough to turn IBM Systems’ red ink to black? We’ll all just have to watch the next few quarterly reports to know.

DancingDinosaur is Alan Radding, a veteran information technology analyst, writer, and ghost-writer. Please follow DancingDinosaur on Twitter, @mainframeblog. See more of his IT writing at and here.


%d bloggers like this: