Back in the day, everything was centralized on mainframes. It was then distributed through client servers, and after a few years was centralized yet again via virtualization. Today there are both hyperconverged infrastructure and disaggregated trends taking over data center architectures. All these changes are happening rapidly to contend with the massive storage challenges businesses face managing data at scale.
Recently, I joined a panel discussion and a webinar with Bo Kennedy, IT Infrastructure Director at OvationData, Gregory Servos, President and Chief Executive Officer of OvationData, and Bruce Gilpin, CEO of Versity Software. Our discussion looked at real-world examples of how businesses can scale without scaling their costs and how cloud service providers and enterprises are solving Big Data challenges with an innovative mix of technologies and infrastructures.
I want to share with you a few takeaways from this roundtable that highlight key storage barriers businesses need to overcome:
1. I’ll Say it Again: Don’t Underestimate the Growth Of Data
With 44 zettabytes of data generation expected in 2020. I don’t think I can overstress how the exponential growth of data and its value (through analytics) is impacting organizations.
We’re not just talking gigabytes or terabytes anymore. Organizations are already storing, accessing and processing petabytes and beyond. Think about it, ten years ago none of us had near a terabyte of data at home. Yet today our PCs, mobile devices, connected homes etc. considered in combination easily and quickly exceed this. From an enterprise point of view things look a lot more drastic. For example, the typical data set of a customer for OvationData averages 400TB, with single files reaching several terabytes in size!
The reality is that companies have to support web-scale applications with billions of objects and geo-distributed data. The ability to store, access and process this data at scale is the lifeblood of any business today.
2. Storage Budgets Are a Strategic Threat
In a recent webinar, Bruce Gilpin of Versity Software, observed that many companies find that if they compare the rate of data growth to their global IT budget, their IT budget will be consumed by storage within the next five years – even with the dropping costs per terabyte of storage. As such, they are under tremendous pressure to find new, innovative and more efficient ways to store and access data. Organizations can’t just keep buying the same brand X to scale out their SAN and NAS systems further and further.
3. Economic Boundaries are the Barriers to Innovation
When you design new data centers, new infrastructures, or new applications, there’s always an economic boundary that’s going to dictate what’s possible. Yet new solutions are trying to lift these barriers and open up new possibilities. This is our goal at Western Digital® – removing economic boundaries to innovation. If we can deliver more, and store more, we can build new applications we never thought possible.
4. You Can’t Run Analytics on Data That’s Sitting on the Shelf
Until just a few years ago, a significant proportion of data was archived on tape and stored in a warehouse for long term retention. But today’s data-driven economy is all about gaining insight from that archived information. Customers want to access their data instantaneously. They want to leverage information on-demand for their own analytics to make better decisions, to meet service level agreements of applications or users, or to open up new services. Organizations can no longer afford to wait for a day, or two, or three until data is ready for processing, compute and analysis. Data can no longer just sit idle on the shelf.
5. Some Tape Stays, Some Tape Goes
If you put data through different technologies, one will obviously be faster than the other, and one will be more cost efficient. Traditionally, tape has always been ahead of disk – certainly not the fastest, but as the most economical solution. However, companies now need data delivered at a much faster rate. Even the fastest tape can take several minutes to load, mount and position before a single byte can be read from it. Then you still need to search through it sequentially. You may find that seeking data from multiple tapes may take hours, if not days.
Until today, disk-based solutions were not large or cost-effective enough for archiving. Yet now with the availability of extreme density drives (like 12TB helium drives) and new object storage solutions, moving data (or parts of it) from tape to disk is no longer daunting.
Don’t get me wrong. I’m not saying tape is dead. However, as customers want to bring their data online, more and more projects are moving off of tape to disk and new ones are launched on disk-based systems. Tape is now used more and more as a backup copy only, rather than an accessible archive.
6. Data Can’t Be Lost
With the value of data growing, digital assets need to be preserved. The data companies store cannot be lost, corrupted or rotted. As such, data protection strategies are changing.
Don’t be fooled into thinking you have to do everything with just one particular technology. This is where an intelligent, yet not too complex technology mix will come in handy. Look beyond managing RAID replication groups to more efficient solutions that use advanced erasure coding. Object storage solutions such as the Active Archive System and ActiveScale™ family can deliver fifteen nines (99.9999999999999%) of data durability. These resilient object storage platforms deliver data durability higher than any others in the industry. Keep your data safe!
7. Simpler is Better
Up ahead is more data, scaling infrastructure, new applications, protocols and workloads. This will inevitably mean complexity. So what you want to avoid is unnecessarily adding to that complexity.
You want to deploy solutions that can scale out and expand with ease. You want the ability to achieve needed performance without being penalized. You want a solution to be efficient, flexible and easy to integrate with other services. What you want to avoid are solutions that require you to buy, implement and manage more products in order to scale up or out as you grow. Keep your stack simple.
8. Compatibility is Your New Best Friend
Companies are taking on hybrid IT strategies – whether cloud, storage mediums or other. Your infrastructure can no longer operate in silos. You want to make sure any solution you deploy is designed to work well with other solutions and won’t lock you in to a single proprietary solution.
Investigate connection protocols (such as new cloud standards, like Amazon S3™) and make sure your technologies integrate well. This will be key in enabling your infrastructure to easily expand to meet your future needs without adding complexity, costs or a complete overhaul. Compatibility and interoperability is your friend.
If you’re looking to learn more about how to break the economic barriers of storage or how companies should evaluate new storage technologies, stream our recent webinar here, or see how to scale to petabytes with ease using the ActiveScale family of solutions.
Too big to store? Stream the on-demand webinar here