Announcing Rackspace MySQL Cloud Database Private Beta

Today we’re pleased to announce the Rackspace MySQL Cloud Database Private Beta for US customers by invitation only.

Overview & Features

Historically, customers running relational databases in the cloud have suffered from poor and inconsistent performance as a result of the limitations inherited from the underlying hardware virtualization platform.  These constraints are common across most infrastructure providers, including Rackspace.

For example, with virtualized disks, there is extra overhead for each operation against the database.  While there have been advances to alleviate some of these constraints, this overhead continues to limit the performance of I/O intensive applications running in the cloud. Other providers have dealt with this limitation by creating custom database storage engines that operate directly in memory.  This approach has great benefits in terms of availability and scalability, but often sacrifices referential integrity or compatibility with your existing applications.

Rackspace has taken a fundamentally different approach to solving this problem by eliminating the overhead of traditional virtualization platforms and utilizing a container-based solution to provide a high performance MySQL relational database service.  This solution provides resource isolation and allows you to quickly and easily automate complex database administration tasks.

Designed specifically for databases, our offering leverages dedicated SAN storage with built-in data replication that gives you high-speed access and redundancy for your data.

Currently, the Rackspace MySQL Cloud Database API operations for the Private Beta include:

• List, add, and remove database instances
• List, add, and remove databases
• List, add, and remove database users
• Enable root user access

However, we are not stopping there.  We are hard at work on a robust set of additional features such as instance and storage scaling, configuration management, backups/restores, and much more.


We are also thrilled to be able to say that we are developing this offering in the open!  Rackspace believes this is the first and only OpenStack affiliated cloud database solution built on top of OpenStack’s Nova code base.  Click here to see our latest OpenStack video with an update from Mark Interrante, VP of Product at Rackspace, and Troy Toman, Director of Engineering for Cloud Compute. Check out and to learn more about these open source initiatives.

Are You Interested In Participating?

As we work through the beta development phase of this project, we’re looking for customers to evaluate and test this service. The ideal candidate will have Cloud Servers currently running transaction heavy applications with MySQL databases in their configuration and be familiar with the Rackspace RESTful web service APIs.

If you would like to apply for free access to the Rackspace Cloud Database API Private Beta program, please take a moment and fill out a brief survey and tell us why you think you would be a good candidate for the beta. Selected participants will start receiving invitations for the Private Beta over the next couple weeks.

We are truly excited to work with our customers to finalize this new offering and your participation and feedback will be crucial to making this product a success.

Be sure to subscribe to our blog for future product announcements.

APRIL 16 UPDATE:   Rackspace Cloud Databases is now available through our Early Access Program.  For more details, click here:


Daniel Morris is the Sr. Product Manager for Platform at Rackspace.

Rack Blogger is our catchall blog byline, subbed in when a Racker author moves on, or used when we publish a guest post. You can email Rack Blogger at


  1. You forgot the words “synergy” and “paradigm”. Honestly, what does “fundamentally different approach to solving this problem by eliminating the overhead of traditional virtualization platforms and utilizing a container-based solution to provide a high performance MySQL relational database service. ” mean?

    Eliminating the overhead of traditional virtualization platforms could be done simply by running a dedicated MySQL server; there, no more overhead of virtualization. That wouldn’t be too fundamentally different. But what is your container-based solution?

    “provides resource isolation and allows you to quickly and easily automate complex database administration tasks.” Some type of quota’s or something? Memory/Hard Drive/CPU quota’s on a running MySQL process on a dedicated server? Automating complex database tasks is usually done by scripting. Will this system eliminate the need for scripts that perform maintenance?

    • Paul,

      You are correct that you can eliminate the overhead of virtualization by running a dedicated box. In fact, Rackspace has solutions to help you do just that. While running on dedicated gear will give you a solid QoS, the hurdle with this approach is that you will have to make a much larger up-front investment in the server and storage gear as well as a higher management burden and longer lead times to get started. By using virtualized systems, you can launch a Cloud Server or Cloud Database in minutes since we only need to start up a virtual machine on already running hardware. Additionally, you pay for only what you need and scale up as demand/traffic increases.

      While the concept of container-based hypervisors are not new, they are not used prevalently in the cloud. Chances are if you are running a database on a cloud-based server, you are running on a hypervisor such as Xen or KVM. While both containers and hardware virtualization are very efficient at separating memory allocations per guest, hardware virtualization is inefficient at isolating processor and disk resources as compared to containers. Database instances rely heavily on memory, compute, and disk to efficiently access data and it makes sense to optimize the infrastructure to support the specific needs of databases.

      With respect to the management features, we currently have the ability to manage instances, databases and users and also have dedicated / redundant storage. We will be working to enhance the features of the product throughout the beta and I recommend you apply for the beta through the provided survey link if you would like to participate and learn more.


      • Hi Daniel. So your Database Cloud product is still running in some type of virtual machine/server, so how does that eliminate the overhead of traditional virtualization platforms?

        Costs aside, would you say a purchased dedicated server running MySQL on it would outperform your cloud solution? The problem with MySQL is there’s a thousand different tweaks and options you can do to it, different database schemas, etc. even the filesystem the database is stored on can affect performance. One tweak here and you make one thing faster but can slow other things down. H

        aving to deal with number of connections, channels, multiple cores, filesystem, threads, etc. becomes a real hassle if you’re looking for maximum performance. In addition, the whole Master/Slave thing is good for simple setups, we’ve even done the Master-Slave-Slave thing, but when the Master dies, promoting a slave to a master becomes a real nightmare, especially when the original master comes back up. So does your solution perform well, better than if we did it ourselves? Does your solution offer easy methods for backing up and promoting master to slave? I’m definitely on the beta list and hope to find some of this out for myself but I was hoping for more technical info on your new setup, which don’t get me wrong, I have very high hopes for. No more dedicated server or Xeround or Amazon RDS solutions for me.

        Also, there’s 2 ways to virtualize Linux. You can do solutions like Xen and KVM, VMware, where you virtualize the entire hardware. The OS is completely unaware that it’s virtual. That means the host system has to emulate a cdrom, hard drive, network card, video card, etc. etc. which does add a decent amount of overhead. There’s also the shared-kernel Virtualization method, like OpenVZ, where multiple Linux machines all share the same kernel, but can be completely isolated from one another. This method uses a lot less resources of the host, is generally much faster and has a lot less overhead.

  2. This sounds really great if its big of a performance increase as you say. I’ve been an RDS user and it is underwhelming. Looking forward to trying this out.

  3. Very good initiative. I would certainly be interested in this functionality if the price is right. Important for me are:
    * Secure/Direct access through MySQL Administrator, preferably through SSL tunneled public/private key access
    * Logging of slow queries
    * Option to log all queries
    * Limit access to database to production server(s)
    * easy backup/restore support


Please enter your comment!
Please enter your name here