Dot Hill improve Framestore’s memory

Storage just got that little bit more interesting, claims Steve Cassidy as he delves into the world of CGI.

Before we get to the nitty gritty, I want to talk to you about Shrek. Neither a Dot Hill nor a Framestore project, but a lovely illustration of the kind of trouble that the movie business can encounter with actually storing and retrieving something quite as immensely vast as the production materials for a fully CGI movie.

There are, in terms of software versions and systems, three different Shreks: Three different animated entities. They cannot appear on the screen at the same time, because the software platforms used to model and render them are utterly different. Such was the pace of development in the CGI business between the three (and a bit) releases of Shrek, that it would be like trying to put a punched paper tape into an SD card slot.

With that in mind, the typically-dull storage business news about Dot Hill and Framestore becomes a bit more interesting. It's not just a matter of stacking disks in a cabinet and watching the blinkenlights. Indeed, the hothouse pressure world of movie production breaks open the usually quite staid and cautious world of storage design and tiering. It really is a matter of commercial competition for these guys, to be able to tweak how many terabytes, at how many hundreds of gigabits per second, they can sling around in the edit suite.

The job in question is all about how many tiers they have in the storage repertoire, and what the speed of a given tier lets them do in terms of keeping recent work close to hand, because apparently the CGI business is becoming a little bit more iterative than it used to be.

That's a nice way of saying that movie directors can be complete divas, when it comes to changing their mind, and you can't predict very readily what a change of mind might trigger in terms of processing cycles for a CGI house. "Make James Bond a bit paler" is (I suspect) fairly low-stress: "make the explosions more intense" on the other hand, probably ties up the whole compute farm for a fortnight.

The fight-back against that kind of problem is all about tiering and the rules for data management. We are used, in more run-of-the-mill businesses, to thinking about fast storage, for things like SQL databases, and maybe a NAS box or two, and then some tapes.

This is, while crude and old-school, a prize example of tiering. You can keep an awful lot of data on tape, but you can only get to it rather slowly: You can reach a lot of data very fast on fast disks, but it will cost you a lot to keep them all spinning, and they are highly unlikely to be earning their keep while that's happening.

Featured Resources

Preparing for AI-enabled cyber attacks

MIT technology review insights

Download now

Cloud storage performance analysis

Storage performance and value of the IONOS cloud Compute Engine

Download now

The Forrester Wave: Top security analytics platforms

The 11 providers that matter most and how they stack up

Download now

Harness data to reinvent your organisation

Build a data strategy for the next wave of cloud innovation

Download now

Most Popular

UK gov considers blocking Nvidia's takeover of Arm
Acquisition

UK gov considers blocking Nvidia's takeover of Arm

4 Aug 2021
RMIT to be first Australian university to implement AWS supercomputing facility
high-performance computing (HPC)

RMIT to be first Australian university to implement AWS supercomputing facility

28 Jul 2021
Preparing for AI-enabled cyber attacks
Whitepaper

Preparing for AI-enabled cyber attacks

22 Jul 2021