You’ve had the NAS vs. SAN debate. You’ve compared the economics of on-premise vs cloud and the combination of both. You may even have bought a shiny new all-flash array for you SQL database. But is your storage “data aware”?
According to Taneja Group, data aware storage is the next big thing, and in a recent webinar, the research form trotted out some startups in this new segment to showcase their offerings.
Senior storage analyst and moderator Jeff Kato said data aware storage is a new category of primary storage system that has several attributes. First and foremost – and obviously – it has increased awareness, which he said means capturing enhanced information and attributes of the content stored on the device. Secondly, it must have real-time data analytics. Thirdly, said Kato, these storage devices must provide advanced data services that act on the advanced information captured.
The last key attribute of a data aware system is open and available APIs which integrate into a larger ecosystem, tools and processes, said Kato. “It’s not enough to have tools internally. It needs to be exposed to the external ecosystem.”
The analyst did make a clarification in that Taneja does not include platforms such as Hadoop or advanced data analytics like data warehouses in the data aware storage category. “These are primary storage devices that will store data and have smarter capabilities.”
So why is data aware storage the new darling? Kato outlined three key factors: the proliferation of compute cores, the ubiquity of flash storage, and software-defined storage, in which software is being created to take advantage of commodity operating systems and hardware components. “These elements are what I see as key enablers to data aware storage.”
Peter Godman, co-Founder and CEO, Qumulo, one of the startups on hand for the online roundtable, said the data storage system itself should be invisible. “If you are aware of it, there’s a problem. The thing that has value is the data itself.” Qumulo was founded three years ago and is addressing what it sees is a necessary shift from managing storage to managing data. And just as the storage should be invisible, he said, all data needs to be visible.
DataGravity, another startup at the online roundtable, is focusing on helping enterprises better understand unstructured data, which comprises the majority of data and is growing at a high rate, said president and co-founder John Joseph. It’s also going after what he called “dark data” – data that has no value unless the user requests it.
Joseph said data aware storage is able to unlock the value of people’s data. “It’s able to identify dormant data, export it and let someone do something with it.” It can also identify if sensitive data is being stored where it shouldn’t be. He said spinning disk and flash are table stakes; the value is understanding data and knowing what can be done with it. “It allows IT administrators to bring a lot more value to the table.”
The advent of companies targeting this data aware storage segment is not dissimilar to that of the rise of flash storage systems – be it all-flash arrays (AFAs) or hybrid-flash arrays (HFAs). Flash, often but not always integrated into flash storage systems in form of SSDs, have gained popularity in the past few years to handle specific workloads where performance is a factor. As the technology has matured and the cost of flash has gone down, flash array adoption has become more widespread.
Research released in early 2015 by IDC concluded that flash arrays are here to stay with adoption growing at a rapid pace, with the worldwide market flash arrays – both AFAs and HFAs – hitting US$11.3 billion in 2014. IDC credits the growth to a wider of variety of offerings from vendors that handle different, increasingly complex workloads as well vendors adding features that enterprises have come to expect on legacy storage systems, with some enterprises now using AFAs for more primary applications.
If data awareness is in fact the next enterprise data storage wave, we can likely expect to see these startups not only provide the features they outlined during the online roundtable, but also enterprise-class data services, including snapshots, clones, encryption, replication, and quality of service. And just like the flash array startups, it’s likely some of these companies will be snapped up by large, incumbent storage vendors who want to bring data awareness to their existing platforms.
