Storage Field Day 6 – Day 2 – Pure Storage

Disclaimer: I recently attended Storage Field Day 6.  My flights, accommodation and other expenses were paid for by Tech Field Day and their sponsors. There is no requirement for me to blog about any of the content presented and I am not compensated in any way for my time at the event.  Some materials presented were discussed under NDA and don’t form part of my blog posts, but could influence future discussions.

For each of the presentations I attended at SFD6, there are a few things I want to include in the post. Firstly, you can see video footage of the Pure Storage presentation here. You can also download my raw notes from the presentation here. Finally, here’s a link to the Pure Storage website that covers some of what they presented.


If you’ve watched any of the Storage Field Day sessions from this year or previous events, you’ll know that the delegates and vendors tend to like to go into technical deep dives on a pretty regular basis. While this also happened with Pure Storage it was refreshing, in my mind at least, to also cover some of the issues that customers are faced with from a business perspective. Pure Storage are keen to make sure that we’re asking the right questions when evaluating the requirements for an all-flash array (AFA) solution. To wit, these are the wrong areas to focus on:

  • Can my array do one meeellion IOPS?
  • Architecture swordplay;
  • Raw price per gigabyte; and
  • How do I adjust the Tiering? RAID? Caching? Disk layout? etc.

Instead, you’re much better off looking at:

  • How does the app perform?
  • What is the usable storage cost for my applications?
  • How simple is this for my admins, DBAs, etc?
  • How do I scale capacity and performance?

I thought Vaughn Stewart did a bang up job of covering off these questions, as well as talking about Pure Storage’s mantra of simplicity.

  • How simple is this for my storage admin?
  • Is storage invisible for my application owner?

There’s no more need for application alignment, it’s invisible. There’s no more need for storage tuning, it’s all about keeping it simple. And automated …

Operate & Automate with:

  • Web-based GUI;
  • CLI; and
  • RESTful API.

And integrated with your favourite stack. There’s support for:

  • OpenStack;
  • VMware VAAI;
  • vCenter plugin;
  • Log Insight plugin; and
  • MSFT’s VSS Provider.

We also need to understand vendor performance claims versus real world requirements. Pure have been looking at a bunch of data and are seeing that the average IO size is 40.6K. As Vaughn said during the session, we need to get out of the “4K Vanity Zone” favoured by vendors when it comes to published benchmark data. What it comes down to is knowing the I/O size of the application you’re running? There’s nothing earth-shattering in this approach, but it’s something we seem to get distracted from at times.

And, I have to say, as much as I enjoy hearing about how storage works in its various incantations, and whether providers dedupe or compress first, and what happens to the data after it hits write cache, sometimes my customers just don’t care about that. And all the technical presentations in the world won’t get them to think otherwise. While I’m not saying you need to forget all about the tech and how it works, sometimes it really is important to keep it simple. I think the Pure Storage message does a pretty good job of that.