All-flash is a synonym for a tough market Last week NetApp launched a new line of all-flash arrays (AFAs) called AFF. It’s based on a specialised version of ONTAP, which has added particular optimizations to read/write data paths. It’s not the notorious FlashRay yet, but it’s not that bad and it has something to say... especially if you are already a NetApp customer. …
Like Mr. Signoretti pointed out, “legacy” is the key word here. All the old players have a huge amount of old code, features and functionalities to maintain in their products. This is a big issue because “old players” often buy new technologies in the form of start-ups instead of developing it on their own. The result is a frankensteinian contraption of different technologies, approaches and protocols which that holds together in an very complex way. New players which have recently designed their products, technology and features are a streamlined business. Very easy to adapt, change and update as needed. This agility means they can easily adapt to new challenges or add new features and technology to keep growing in a very competitive market.
Pure and Solidfire have concentrated on offering AFA block storage because applications are the biggest performance demanders in the storage ecosystem. This made concentrating on them the natural choice for a start-up because for these companies focus is mandatory. This is not to say they won’t move to offer more in the future.
The conclusion is this: new players in the storage market can achieve expected market growth because their agile infrastructure will allow them to add more functionality a lot easier. They have the experience of developing technology and integrating it in a streamlined, quality product.
This post has been deleted by its author
This article captures key trends that are becoming evident to everyone: Flash technology is here to stay. Tiered storage systems based on hybrid arrays deliver cost effective and performance optimized storage.
What makes the real difference is how data is managed and moved to the appropriate tier. With today’s workloads driven by people accessing information in real time, and machine data that is collected and communicated continuously, the resultant workload on storage is fundamental different than in the recent past. Some projections anticipate 50 billion devices connected to the Internet by 2020, all gathering and sharing information in real time. So, what will differentiate solutions from vendors is how hybrid storage systems identify and move data among tiers. Look for those that do this in real time. Real-time tiering makes the difference.
I applaud Enrico's recent article in the Register. Particularly when he states, "they've got to have products capable of serving blocks as well as file or objects, and build integrations (or integrated data services) on top of that." Why should customers be forced to choose? What is needed is one unified storage architecture that cost-effectively supports block, file, and object. Only 4% penetration for all-flash arrays means we are not yet meeting the customers’ needs. Effective unified storage is possible and should be everyone’s focus.
Gareth Taube, VP Marketing Infinidat
... and NetApp gets it. Pure, SolidFire and Kaminario also get it ... they just can't deliver it like NetApp can.
They are shiny baubles ... and the sparkle only lasts for fleeting moments in a fast moving industry. NetApp may not have the sparkle, and the paint is original, but those are often the most sought after and valued creations.
We shall see...
Flash media is here to stay. The cost of Flash media is declining rapidly and in order to gain the benefit of Flash media and other new technologies the modern AFA players built a new architecture that is the next generation storage architecture. A modern AFA architecture like Kaminario is a true agile and dynamic infrastructure that meets new customer challenges. It is a scalable and cost-efficient architecture and today it is concentrated on solving real applications needs. This is why the new AFA architectures are winning when competing against the legacy architectures.
I totally agree that storage will continue to evolve and these system will need to add more functionality in order to continue with the exponential growth. Here comes another advantage of the new modern architectures. The new modern AFAs built a true agile infrastructure that will allow them to add new functionality a lot easier compare to legacy architectures that were not designed for the new challenges. I expect to see the AFA vendors adding new modern functionalities faster and by doing this continue to have products that are a better fit to new evolving customer needs.
Shachar Fienblit, CTO, Kaminario
NetApp Employee Opinions are my own.
@Shachar So your thesis is this ...
Flash = Next Generation = Good
Startups = Modern = Clean Code Base = Promises of fast development = Good
Existing Vendors = Legacy = Crufty Code Base = Slow addition of new functionality = Bad
With nothing substantive to back this up.
I could easily counter with
Flash in external Arrays = Sustaining innovation = Good
Startups = Focussed on Time to Market = Buggy Minimal Viable Product = Roadmap Marketing = Bad
Existing Vendors = Solid QA, Engineering and Support Infrastructures = Proven Cadence of Innovation = Good
But without specifics and proof points, its just more tech marketing fluff that adds nothing to the conversation. There is some truth in both positions, some of the larger established vendors haven't done much innovation in their core product lines, and some startups are on the verge of implosion as the technical debt from hurried development kills their ability to execute.
I can't make substantive comments about the state of Kaminaro or any other startups code-base, but I can say that the changes in ONTAP that enable All Flash FAS is just one of the innovations NetApp has developed recently, theres plenty more (vVol support, QoS, Automated non-disruptive upgrades, enhanced application support, SMB3, NFS 4.1, Cloud integration, policy defined transportable multi-tenancy containers, Cisco ACI support etc, etc)
IT people are smart, they've seen simplistic "Four Legs Good, Two Legs Bad" FUD before, and from my experience they all see through it.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
