Interview: Spectra Logic on Big Data Storage in the Cloud

Print Friendly, PDF & Email

Big Data has transformed the way organizations view data and what can be gleaned from it. The cloud has also changed the way organizations see data and what can be done with it as well as how it is stored. Spectra Logic has a suite of products designed to provide solutions for Big Data, cloud computing, and more long-term storage. We caught up with Kevin Dudak, Spectra Logic’s Senior Product Manager, to get further insight.

insideBIGDATA: As mountains of data continue to amass, the need for robust storage solutions is apparent. What does Spectra Logic do in this space to set itself apart?

Kevin Dudak

Kevin Dudak: The unrelenting growth of data has clearly impacted the way we think about data storage. Tape storage technologies have evolved substantially the past five to ten years. At Spectra, we’ve worked tirelessly to lead innovation and challenge expectations of the data protection market with intelligent, integrated, and simple to use backup and archive technologies. Our suite of T-Series tape libraries range from high capacity Enterprise class libraries, such as the T-Finity and T950, to smaller SME-focused libraries, such as the T120 and the T50e. We offer high capacity TS1140 and open standard LTO media options, have the capability to offer block, file and object storage on our tape systems; and deliver long-term storage for under $0.10/GB LIST pricing.

nTier Verde, our 4th generation disk product, takes a different approach in the disk world.  Where most disk systems continue to add features, capabilities and the associated complexity, Verde only has what is needed for large data storage.  By keeping it simple, affordable and by focusing on data reliability, Spectra provides a storage system that meets growing needs.  Combining new interfaces and products with a continual evolution of core capabilities, such as data integrity, is where we excel and continue to set ourselves apart.  With the example of data integrity, all Spectra products have integrated data integrity features and scans.  This is particularly important as data sets grow to quantities many people never imaged.

insideBIGDATA: Who traditionally are your customers? Which verticals does Spectra go after?

Kevin Dudak: Our traditional customers are a mix of IT backup and organizations with large data stores.  Some of the verticals that have generated large amounts of data are Media & Entertainment, High Performance Computing (HPC) and Energy.  The interesting thing about these verticals is not only have they traditionally created a lot more data than the broader technology market, but they have always focused on the significant value in their data.  The demands of these markets have helped position Spectra well for the Big Data tsunami being seen today in the broader IT market.

insideBIGDATA: What has Spectra done to keep pace with the proliferation of storage in private, public and hybrid clouds?

Kevin Dudak: Spectra has been involved with various cloud deployments for years.  In fact, we had several customers that had cloud offerings based on Spectra tape libraries before people started calling it a “cloud.”   Today, our storage systems are installed in public, private and hybrid clouds.  When looking at cloud needs, we continue to look at ways to address the needs of scale and operational costs.  Public clouds need systems that are not only very cost effective, but have a predictable cost curve.  As we update existing products, and release new ones, you see new capabilities that specifically make it easy to tie to different type cloud deployments, such as our DS3 interface to Deep Storage.

insideBIGDATA: What are the advantages of cloud solutions? Limitations?

Kevin Dudak: Cloud storage opens up a lot of new and interesting opportunities.   Cloud is great for dispersed content access and distribution.  If you have data that lots of people need to access from many different locations, cloud is hard to beat.  For smaller organizations, and start-ups, the lack of capital expense to get started can be very helpful.  Just write a small check every month, and you can store your data.  A cloud can make a great off site location for disaster recovery (DR) as well.  However, there are some limitations with clouds that should be considered.  Often, you do not have control over where your data is stored, this can be an issue for some data types, such as medical records that may have regulations on where the data resides.  Many cloud providers offer the options of multiple copies in multiple sites as a way to add redundancy.  While this is helpful, most providers run the same software and hardware on all locations.  A simple firmware bug could affect all of the data.  While clouds are very flexible, they can make it hard to truly control your data.  For instance, can you guarantee that the data was deleted?  If you can’t this could be a problem for data sets that require confirmed deletion.

insideBIGDATA: What advice can you give to those considering a cloud-based storage solution?

Kevin Dudak: There are a couple of key things to think about.  First, how confident are you in the cloud provider’s long-term viability?  Nirvanix was only able to give customers 2 weeks’ notice that they were shutting down.  If you have a copy of the data outside your cloud provider, you needed to find another service provider and upload your data.  If Nirvanix stored the only data copy, you needed to retrieve the data and store it elsewhere before they shut down.  Do you have the capacity and bandwidth to get the data back within a short time frame?  You should also look at the long term price of the cloud solution.  For smaller data sets, cloud can be very attractive.  However, for large data sets cloud storage may not be cost-effective.  Plus, you will pay for the first GB of data you write to the cloud today and every month 10 years from now.  If you access a lot of data, then the costs typically accelerate faster.  Very large data sets are almost always more economical to store in onsite storage versus a cloud.

insideBIGDATA: Spectra has been around for over 30 years. To what do you attribute this longevity?

Kevin Dudak: Spectra’s longevity can be attributed to a number of things.  It starts with our focus on our corner of the storage market.  By having a laser focus on data protection and preservation, we are able to see needs and opportunities that companies with a broader focus miss.  Our passion as a company is innovation.  The rock stars of our company are the engineers and architects that constantly drive our products forward.  We are able to bring it all together in ways others struggle to do in part because we are focused and agile, plus our team is located in a single campus.  Support and engineering sit next to each other, and manufacturing is in the building next door.  Having everyone together helps to increase efficiency, and makes it easy for us to make changes quickly if needed.

insideBIGDATA: You have weathered the changes in data storage over the years well, so you must be looking to the future. What trends to do foresee for data storage and Big Data environments, and what does that mean for Spectra?

Kevin Dudak: Change has been consistent in data storage over the last decade. We are seeing evolutionary changes within data storage that are both ‘storage centric’ and ‘data centric.’  Big Data is an excellent example, showing how unexpected information and value can be obtained from what might have been valueless data.  It is truly changing how people look at and value their data.

Big Data puts new requirements on storage with respect to scalability, data integrity and cost efficiency – and Spectra is well-positioned to serve this market. Our archive and backup data storage tape products support all aspects of secondary storage, are compatible with every major tape and disk format, enable massive scalability and provide a plethora of advanced features that ensure the data is protected, its integrity is maintained and that it will be available virtually forever. Our suite of T-Series tape libraries offer high capacity TS1140 and open standard LTO media options, have the capability to offer block, file and object storage on our tape systems; and can deliver long-term storage for under $0.10/GB LIST pricing.

We are also leading the market in innovation by providing new ways to access and distribute data, such as RESTful interfaces and extremely low-cost storage for public, private and hybrid clouds. Our focus on innovation leadership allows us to help organizations implement the best storage solutions for their specific needs.

Speak Your Mind