Scheduled Images Available Now On The Open Cloud

One of the recommendations we often make to our Cloud Servers customers is to always ensure you have valid backups of your servers and data. One of the most critical ways you can ensure this is to take a regular “image” of your Cloud Server. Today we offer you access to Scheduled Images for your next generation Cloud Servers.

Note: We often use the term “image” to represent the base images from which Cloud Servers are built, and the term “snapshot” to refer to a point in time image taken of a running Cloud Server. For all intents and purposes the two terms are interchangeable; however, because the feature is called Scheduled Images, this post, the online documentation and the Control Panel have standardized on the term “image” or “server image.”

Users of our first generation Cloud Servers have long had the ability to setup a fairly fine-grained schedule of daily, weekly and monthly rotating server images. When a point in the schedule is met, an image is taken of the server without causing downtime and stored in your account for recovery or rollback purposes later.

While this system offered a great deal of flexibility, it also introduced complexity, and put our users in a position of ensuring that they designed valid, non-overlapping, non-conflicting schedules all to perform what is ultimately a straightforward task.

When we designed Scheduled Images for next generation Cloud Servers, we started with simplicity in mind: require the fewest number of choices to ensure that more customers are able to setup Scheduled Images more easily, and thus have more protection in the event that they need to utilize an stored image.

So, instead of a traditional schedule that combines times and daysarbitrary retention periods and overlapping rules, we focused on the one configuration item that our users found most important, and that’s the only question you have to answer to get up and running:

“How many images do I want to keep of my server?”

Scheduled Images in next generation Cloud Servers relies on that one simple question. You provide the number of images you want to retain, and we will take one daily snapshot / image of your server and store that number. It as easy as that. If you specify seven images, then when the eighth image is taken, the oldest will automatically be removed from your account.

Besides highly simplified scheduling and retention, we’ve also improved the capabilities of our first generation Scheduled Images in a number of areas.

Perhaps of the biggest change is that there is no longer a size limit on your server when taking images. This means that you can now take an image of a 512MB Cloud Server, or a 30GB Cloud Server, whatever you need.

Now, physics being immutable (so far), of course the larger the server, the more data will be stored, and must be transported across our internal networks into your account. We’ve done a lot of optimization to try and ensure that no matter what size your server is, you’ll see a successful snapshot for as many days as you’ve chosen. However, there may be scenarios where a very large Cloud Server takes longer to image than expected, or other limiting factors. If this occurs, no image will be stored for the day, and you will not be charged. Cloud Servers will re-attempt the snapshot the next day, without requiring any user intervention.

We hope that you find the new Scheduled Images feature useful. We’ll be working with you and our partners to continue to evolve Scheduled Images, and we absolutely welcome your feedback.

For more information, check out the Scheduled Images FAQ in the Rackspace Knowledge Center.

Richard Goodwin joined Rackspace as a Product Manager in June 2012, focusing on images and the OpenStack Glance project in the Cloud Compute team. Before joining Rackspace, he helped manage and launch a series of online services, including online backup, in the group at Symantec Corp. He has a long history in both web services and storage technologies, with almost a decade each at Symantec and Dell Inc. serving enterprise and SMB customers.


  1. If a scheduled image fails, is the customer notified? What if the image/backup fails several days in a row?

    If a customer wants more images per day ( think in the case of a database where having the most recent records possible is important ) do customers still have the ability to schedule in a more granular way via the API?

    We have specific retention policies where we keep 2 weeks of daily backups, 5 weeks of weekly, and 12 months of monthly. How would we achieve that in this scenario?

    • Angela,

      A couple of answers here:

      1) We are looking into ways to improve the ways you can be notified if an image fails. Right now, it will simply retry at the next available time, so you shouldn’t have ongoing failures.

      2) Currently, a single image per day is the maximum allowed via the schedule function; however, you can certainly use the API to call and create a manual image at different times of the day.

      3) Again, currently the functionality is meant to be simple to use and configure, and meet the needs of common recovery scenarios (typically rolling back within a few days), versus being a complete backup solution. For both #2 and #3, we strongly recommend using a backup solution like Cloud Backup, that can have targeted time frames, more granular scheduling, and is more effective for backing up specific data vs. the entire server image.

      I hope that helps answer the questions; we do know there are users with more complex needs, and we designed the feature for future expandability, so keep the the feedback coming. We’ll continue to listen and tune things up!

      Cloud Servers Product Manager


Please enter your comment!
Please enter your name here