IBM executives came to Boston earlier this year to unveil the company’s Elastic Storage solution, Big Blue’s most significant step into the increasingly competitive software-defined storage realm.
The Elastic Storage announcement was significant enough that top officials—including Tom Rosamilia, senior vice president of IBM’s Systems and Technology Group and integrated supply chain business, and Jamie Thomas, general manager of storage and software-defined systems at IBM—headlined the event, which also featured several partners and customers talking about the virtues of the solution.
IBM also introduced Elastic Storage a week after EMC unveiled its ViPR 2.0 technology, helping give some shape to the burgeoning software-defined storage (SDS) market.
However, IBM also used the May 12 event—which also was Webcast—to highlight some of the work researchers at IBM Labs were doing around storage. According to some IBM officials, the announcement earlier this year of the deal to sell the company’s x86 server business to Lenovo for $2.3 billion caused some confusion in the media and among customers about what IBM was doing with its storage business.
“There have been questions about IBM’s commitment to storage and its importance to us,” Thomas told eWEEK.
To answer those questions, several IBM researchers were on a panel discussion at the event in Boston discussing some of the wide-ranging work IBM Labs was doing in storage technology. One of the projects discussed in some detail involved the development of “storlets,” a research prototype that is built atop OpenStack software and designed to make managing object storage easier and more flexible.
Object storage has been around for several years, but Michael Factor, Distinguished Engineer in research at IBM Labs in Haifi, Israel, said managing such storage can be difficult and frustrating. With object storage, data is stored as objects, which contain not only the data—such as documents, movies and images—but also the associated metadata that does such jobs as describes the data’s content, who created the data, how the object is related to other objects and how the data should be managed.
IBM researchers are using software-defined principles to develop technology—storlets—that will make it easier for customers to more quickly and dynamically get value from the objects. For example, through the storlets, users would be able to extract greater detail in images. If a doctor wanted to get information from a medical image on his tablet, rather than having the entire image downloaded, the storlets software would enable him to extract only the part of the image that he needed and have it sent to the device.
A key to the storlets software is that the stored data can be processed locally, rather than having to be transferred from the storage server via the network to another computer to be processed and then sent back to the storage server, Factor said during a talk at the Boston event.
“We bring the computation to the data, instead of bringing the data to the computation,” he said.
IBM Looking to ‘Storlets’ to Manage Object Storage
In a post on the IBM Research blog, Factor compared the process to preparing a meal.
“Imagine if every time you wanted to cook a meal, you had to bring all your ingredients to a central neighborhood depot where stoves, appliances, and cooking utensils were available to ‘process’ the food,” he wrote. “That’s similar to the current situation with data on the cloud. Storlets come to remedy this situation by moving the heavy lifting to where it’s needed—similar to allowing you to cook everything in your kitchen, where all your raw materials are already located.”
What this means is that users can get the data that they want more quickly, network latencies are reduced, and bandwidth is saved by not having to move the data over the network to be processed.
“Our vision is to reduce costs, increase flexibility and improve security by turning the object store into a platform, and allowing the functionality of the object store to be extended using software,” Factor wrote.
In addition, the storlets could be a way for businesses to offer new services, such as image alignment storlets or image extraction storlets. The software can extract the metadata—such as size, subject and format—from each object. With these capabilities, users gain greater control over the data. Building on the doctor and the medical image, users could take a movie uploaded to an object store and have the storlets automatically create a single image, or a have documents sent in a specific format.
“All this sophisticated computation is moved into the storage infrastructure via the software of dynamically loaded storlets, making it faster, more flexible, and far less expensive,” Factor wrote.
Storlets leverage OpenStack Swift and work with Hadoop and the OpenStack Cinder block storage service, and will be able to be used either in the cloud via IBM’s SoftLayer or on-premises, he said. IBM researchers are taking advantage of projects in the European Union to help develop storlets.