Finally – from brandbox to whitebox: Storage fabric is SDS realised A new term is popping up among startups: the storage fabric. Depending on who you talk to, the term itself is recycled, but the concepts and implementations definitely are new. Today's storage fabrics are closer to the originally hyped promise of Software Defined Storage (SDS). What remains to be seen is if the arrival is timely …
While I welcome our storage fabric overlords, I doubt that it will render hardware choices totally irrelevant. Sticking to the HCL is very much required in the VMware world, almost equally true in the Microsoft one. Meet our newest abstraction layer. Should be an interesting ride, for certain definitions of interesting.
It's worth pointing out that fabrics can install onto bare metal by integrating with a Linux kernel. So HCLs can be the Microsoft HCL or the VMware HCL if you ware using the hyperconverged VSA-style solution on top of an existing hypervisor, or it can be "pretty much anything you can find" if you are using a bare metal scale-out solution native to the fabric vendor's software/integrated with KVM. Interesting times.
Trevor, this is a good article and would have been even better two or three years ago. Unfortunately the "big players" already got there in terms of enterprise level virtualization of white box storage, ingestion of existing storage and server based SAN & NAS. Take a look at Dell EMC ViPR currently running version 3.x and scaleIO pretty much the same. Massively scalable technologies already in enterprise customers delivering what you are talking about with a level of maturity that startups cannot match.
The software defined component is important because you still require the means to control the "whitebox" technology to deliver the services that you need. To do this you need a virtual array to mimic the data services of a fully functional physical storage array, particularly in the NAS arena. The virtual array also offers QoS, encryption and data services like tiering and compression.
As I said nice article, but, it is not a view of the future but of current reality.
ViPR's cute, but to my understanding is still pretty limited compared to modern SDS platforms. Yes, proper fabrics exist today, but none of them have all the pieces together and have enough funding to go big. There are startups with all the bits, but they're tiny. There are big companies with SDS solutions, but they don't have all the bits.
Storage fabrics are today where hyperconvergence was circa 2010 or 2011.
Also: ScaleIO as a storage fabric? No. Hyperconvergence is only one feature of a storage fabric. It is not a fabric in and of itself.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
