Recap: Rackspace Solve SF 2014

Update on the Floppy Drives:

Rackspace’s @waynewalls: @scottefein, if you’d like a floppy drive to perform an Office Space penny skimming scheme at Initech, we could probably help :)

I had the pleasure today of attending Rackspace Solve, a new Rackspace conference focused on showcasing Rackspace success stories. While arguably incredibly self-promoting, I learned a lot about Rackspace’s services, things I wasn’t aware of at all. Now WePay uses Rackspace’s dedicated infrastructure, but I’ve had a lot of experience with Google Compute Engine, Amazon Web Services, and DigitalOcean.

Here’s where I started: “Rackspace lets you spin up cloud servers, similar to AWS, just with actual support”.  I can attest that Rackspace really does have fanatical support, and for certain (single server) applications I used Rackspace for that reason.

But what happens when you’re running more than one server? Or you have a big infrastructure spend? Turns out there’s a whole new world..

With Rackspace Hybrid Solutions, you can run dedicated servers and cloud (as-needed) servers in the same network, served by the same load balancers. Essentially your cloud servers and dedicated servers don’t have to be different-you can treat them exactly the same. Now that’s pretty cool.

When I say dedicated servers, I don’t mean AWS dedicated infrastructure-where you have a 12-36 month contract in exchange for a cheaper rate, or Google Compute where you just use it a lot and get cheaper rates. I mean they actually can build servers to your specifications (takes 5-10 business days to add a server[UPDATE: If you choose from their prebuilt catalog, they can get a new one on in hours]).  Want 10 TB SSDs? Check. 20GBs of RAM? Why not. Want a floppy drive? I assume you could ask for this, but if you need a floppy I question your business model. Anyone from Rackspace want to comment if this? (Turns out YES)

Ever run a database in AWS and need to reattach your storage to a new instance when your server has issues?  Yeah, you either need to programmatically figure out how to identify this or have a manual procedure in place.  Turns out in Rackspace Dedicated, it will actually just switch to a new server when it fails automatically. As long as your server can handle a reboot (and if it can’t, it definitely should) you’re looking at maybe 20 seconds of your DB server being down.  That’s a lot better than the alternative…

Do you like geo-based routing? Well they can hook you up with GeoTMS to do a load-balancer based geo-based routing via an F5 load balancer. It’s super cool, you need to check it out. In the past I’ve used Route53 for this, but this is a real, grown-up solution.

Sorry for spending this post fanning over Rackspace. But I was really impressed with their offering. And we’re already a customer haha.

Don’t count Rackspace out when you’re picking your next hosting provider… Best advice-call them. Tell them what you need. You’ll be pleasantly surprised

Leave a Reply

Your email address will not be published. Required fields are marked *