• Sales: 1-800-961-2888
  • Support: 1-800-961-4454

Rebuilding a Server


Rebuilding a server replaces all of the data on an active server with either a different, new image or a previously saved image of this same server. The new rebuilt server will maintain the same IP address. This article describes how to rebuild a server.

To rebuild a server using the New Cloud Control Panel, follow these steps:

    1. Open the New Cloud Control Panel, and click Servers at the top of the page.

    2. Click the Actions cog next to the server you want to rebuild and select Rebuild from Image

      Resize Server in the Action Cog

      A pop-over window appears with the Rackspace and Saved tabs.

      Image Tabs

  • To rebuild your server with a new image, select an image from the Rackspace tab.
  • To rebuild your server from a previously saved image, click the Saved tab.
  1. Select an image and click the Rebuild Server button.

The server status will change to Rebuilding and will be unavailable during the rebuild process.

Related Information

Learn More About Cloud Servers



© 2011-2013 Rackspace US, Inc.

Except where otherwise noted, content on this site is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License


See license specifics and DISCLAIMER

15 Comments

The above procedure works fine for linux, but for windows, although the server comes up, you can't get into it with a remote desktop connection. This is because the server invariably comes up on a new IP address, and this IP address is not set in the windows configuration as part of the build process, annoyingly.

The only solution is to go in with the console, go into the network settings and set up the external IP address that Rackspace has allocated to the server. The Rackspace console option on the website is a VNC mechanism that requires you to have Java installed - which has recently been removed from all our desktop PCs in the company due to the continual security issues using it.

This was not a problem initially when we started using Rackspace, but has existed as a problem for at least 3 months now - it may be an OpenCloud issue. If someone at Rackspace is reading this, can they look at this problem - at the moment I'm having to power up an old PC that still has java on, and go through the laborious process of using the java console. It ensures that although the server is created in a few minutes, its an hour or so after before I can actually use it.

Interesting. I will pass this information back and see if we can find out what's going on. Thanks, Martin.

Has this been fixed? I need to switch from Linux to Windows and need to know it won't be a problem?

It is my understanding that the issues Martin experienced are fixed in our latest Windows Server images.

As of today, this is still a problem.

Thanks James. Would you mind opening a ticket for the issue so our techs can look into it? If you send an email to kc@rackspace.com with the ticket number I can make sure the ticket gets seen by the people working the problem.

I can confirm that for a 2nd generation server, this problem appears to no longer exist. I have created a new Windows 2008 server, stored the image and deleted the server, then recreated it from the image. It was, as expected, given a new IP address, but this time that IP address was correctly set up in the windows network configuration and I could immediately remote desktop into the server.

Unfortunately, there is no way of migrating a first-generation windows server into a 2nd generation, so until I completely reinstall a new server with the same applications as the old one and migrate to that, I will continue to have the problem.

What is the process of creating a cloud server from an image - is it all automated? I'm trying to understand why it takes so long and why it hangs at 17% for most of that time. Is it possible to provide better feedback than 17% (which jumps to 50-something%, etc)? Maybe listing the actual task or something?

When you're trying to add capacity on the fly, there's nothing as frustrating as watching it sit at 17% for 20+ minutes and no other feedback is available.

The process is automated. What happens behind the scenes depends on whether it's a first-generation or next-generation server, but the progress reporting could be improved for either environment. I know that work is being done on better integration of OpenStack Glance (the OpenStack imaging system), so hopefully improvements in progress tracking would accompany that change.

Improvement would be welcome! It's a next-gen server. It was 17% complete for 26 minutes and has been 83% complete for 41 minutes and counting. It's a great idea, but not the promise of rapid scaling that the "cloud" is supposed to offer.

It would be less frustrating if there was better communication of what was actually happening. Percentages are useless when they make huge jumps and the length of time at each step is so long.

I can't argue with that. You might also post a suggestion on our product feedback forum to make sure your suggestion is seen by the right eyes:

http://feedback.rackspace.com/forums/71021-product-feedback

I can say the build speed will be influenced by both the size of the image and the number of files in the image. Removing a lot of temporary and session files from the server before taking the snapshot can make for faster builds from the resulting image.

The percentages are rather misleading. "Step 1 of 6" would be way more helpful than "17%."

I experienced the same thing happening when I attempted to rebuild a server from an image, I could only get onto the server from my control panel and had to log out of my browser whilst java updated, when it had only just done that update about two hours ago. My support did not help me through this process, and did not mention it as a possible problem.

Thanks Edith. I'm talking to our devs about this, and hopefully we can come up with a solution to the issue for older server images. It's looking like the problem can be worked around for now for most images by ensuring there are no pending critical updates on the image, but there are cases where the problem continues to happen anyway.

Same thing just happened to me (GEN1 server), and this is how we fixed it. Use the (terrible) Java based VNC viewer here in the control panel to access the server. Then you do this:

Control Panel -> Network and Sharing Center -> Diagnose and Repair

After it's done your server can be accessed and it can also access the Internet itself.

Hope that helps somebody, and thanks to the RackSpace tech helping me - I forgot his name already (it happens when you've been awake for nearly 24 hours)

Thanks
-pb

Add new comment