5 min read
Editor's Note: Watch the complimentary webinar.
There’s a line in Alice’s Adventures in Wonderland that says, “It's no use going back to yesterday, because I was a different person then.” Some days, I feel that way about my storage infrastructures as well. I fell down a particular rabbit-hole into storage management accidentally almost 20 years ago.
The typical application I supported was a database platform. With a monolithic app running on one very large server, we obviously used one very large storage array that, while slightly more modular, still had its configuration limitations. If you outgrew your server, you bought the new, bigger version. If you outgrew your storage, you either added another rack or migrated to a bigger platform. That’s great for forklift rental firms but not so great for admins trying to scale granularly or modularly (or interactive-modularly).
A little over ten years ago, applications started to scale out in force, often using distributed and replicated storage in place of monolithic storage. Grid computing, HPC, and Hadoop were the most evident forms of this scale-out server and storage model. But applications either had to be rewritten from scratch to take advantage of the new paradigms, or new applications were written to layer on top of the traditional database platforms and storage. Either way, it wasn't a very smooth transition, unless you had a development team, a noticeable budget, and a lot of time to work on the process.
In short, re-platforming became possible, but it remained painful at best, and disastrous at worst. I've been through both scenarios in my career.
Here's where MapR comes into play. If you're reading this blog, you probably know about the MapR Data Platform, and odds are you know that MapR has had multiprotocol storage access in its platform for several years now. Centralized management of the platform features, including numerous APIs for storage access, simplifies the operation of a MapR cluster or clusters on the software level.
Where does Cisco come into the picture? With the Unified Computing System (UCS), Cisco brings uniform management and monitoring to server platforms across blades and rack-mount servers, supporting a wide range of storage from standard SATA and SAS drives through SSD, PCIe flash, NVMe, and whatever comes next. Through the use of service profile templates and a scalable, repeatable configuration from the bare metal up, you have the same central, sustainable management for the network, server, and storage that you do for the software from MapR.
With hardware and software converging, and making itself a platform accessible to legacy applications as well as modern data systems, we find an opportunity to integrate those applications on a storage platform that scales by the server, not the shelf or the rack or the cage. We can integrate different performance requirements, grow as the use cases require (not only as a vendor wants us to buy), and make our data more accessible in real time as well as for the long term.
In Alice’s Adventures in Wonderland, Alice asks the Cheshire Cat,“Would you tell me, please, which way I ought to go from here?" and he responds: "That depends a good deal on where you want to get to."
My friend Bill Peterson, MapR's VP of Partner Strategy, and I will have a conversation on Wednesday, January 25, 2017, about where you might want to get to and which way you ought to go from here.
We'll talk about putting your data and application re-platforming on fast-forward with Cisco and MapR technologies. We'll discuss how data and applications have changed, what our two companies offer to make the most of your data (including the MapR CDP and Cisco's C-Series and new S-Series server offerings), and even cover when/why you might choose not to move your use cases to our CDP on UCS platform.
Register above, and we'll see you soon. If you can't make it on Wednesday, register anyway, and we'll have an on-demand version available soon after the event.
Stay ahead of the bleeding edge...get the best of Big Data in your inbox.