Wanting to understand why hardware and SDS are independent, and yet why they are linked? Why not everyone is best served with a “DIY”model to SDS? (BTW – customers I talk to using SDS + COTS at scale are all learning this).
This will all be very germane in the coming two weeks. Read carefully.
Fascinating research paper, linked below (you will need to buy the paper, but if you’re thinking about it – the cost of the paper is fractional compared to what you will learn, and on 2/16 I will share a personal story).
Not an ounce of emotion from me on the topic of SDS + SW, SDS + validated hardware, appliances from me – every customer is different. We have fully embraced SDS (VSAN, ScaleIO, ECS, Isilon SD Edge, DD/VE, Recoverpoint for VMs, etc, etc) and all the consumption form factors (software only, software + hardware, full appliance). Rather – this is a case of “nice to see someone else do the full math”.
This is a 3rd party (Gartner) echoing the “SDS SW only –> Bundled with validated hardware –> appliance” examples I’ve been echoing for a while here, here, here (based on a ton of customer interaction).
How about the cost of Cloud Storage models, compared to “DIY”, and storage appliances? Again, good third party work. This is a case which is pretty analytical, and while there’s a lot of opinions, the math holds up pretty clearly. Again, not free (I wish you would Gartner!), but worth the $$ if you’re debating this choice.
There are a TON of opinions on this topic, these two represent some of the clearest mathematical assessments of the economic and non-economic considerations inside this (relatively) new domain of different consumption and assembly models of the IT domain of “persistence”.
Pure SDS is here to stay – it makes a ton of sense in many ways, and it gives a certain economic and architectural transparency and flexibility. BUT – customers need to decide where they are on the “buy vs. build” continuum.
This links to a lot that is happening in the next couple weeks – stay tuned dear readers!