2

One of our Windows Server 2008 servers which was hosted as a Cloud Server on Rackspace has recently failed. We are not getting much help from the Rackspace support, all they keep saying is that hardware on the physical server has failed and data on the VM server corrupted.

Did anyone had such an experience in dealing with Rackspace?

What are our options to retrieve the data from the corrupted VM?

I know they running VMs on Xen platform and if I had that VM locally most likely I could retrieve the data even if Windows is corrupted.

This server was running our main website and DB, any help would be greatly appreciated!

Elvin R.
  • 123
  • 2
  • 2
    Has anyone else had this issue with Rackspace? Probably. It happens to any provider. Hosting in the cloud isn't really magic and doesn't automatically provide 100% uptime and reliability; this is why you need to engineer your solution for redundancy and include backups as an integral part of your business. If you make money off the machines hosted, you need to take into account disaster recovery! – Bart Silverstrim Jul 22 '12 at 02:24
  • 1
    I work for Rackspace, and I'd like to look into your case for you. If you don't already have a ticket created for his issue, please create one through the control panel and email it to my team at help@rackspace.com. I sincerely apologize for any disruption this issue has caused you. Best, Jeremy Wasner Rackspace Hosting – Jeremy Wasner Jul 22 '12 at 02:17

2 Answers2

2

This question is not going to really be answerable by anyone other than Rackspace, but in the general case:

If you have a hardware failure that causes data loss or corruption, there are three basic options available to you.

  1. Restore from a backup. This is generally the best option. It's the cheapest, fastest way to get your data back and your system running again. Of course, it does require having working backups, which doesn't sound like the case here.
  2. Data Recovery. This is the next option, and is not generally preferred because it's resource intensive, takes a long time, and may not be able to recover what you need anyway. This can take many forms, from something as simple as using "the freezer trick" on a failed hard drive and copying the data off, or as complicated as actually trying to undo the corruption by analyzing the raw data on the disk. Or of course, you could hire someone to do this for you, but professional data recovery is time-consuming and very expensive.
  3. Live without the data. Probably the least palatable option, but it's what happens from time to time when you don't back up your important data.

In your case, since the hardware isn't yours (ruling out 2), and you don't seem to have backups (ruling out 1), it looks like you're left with option 3. I'd look at my ToS/contract, and scream bloody murder at the cloud provider that didn't have actual backups of my data (as I'd expect), but if it's not in the contract or agreement you signed that they'd do backups and protect you from data loss/corruption, you don't have much recourse.

HopelessN00b
  • 53,385
  • 32
  • 133
  • 208
  • 1
    You'd expect a cloud provider to have backups of your data? Good luck with that one. – womble Jul 22 '12 at 01:59
  • I would, but I'd also expect the terms to be written out and agreed upon before moving anything to "the cloud." I've worked with some local DC providers who actually offer this type of service (for an additional cost), and haven't had any problems with it yet. Not my preferred approach, but when left with no other options, I jump at the opportunity. – HopelessN00b Jul 22 '12 at 02:05
  • Sure, we do managed backups too, but we're not a "cloud" provider -- I'm yet to see a true "cloud" provider offer backups as an *integral* and indivisible part of the service, and quite honestly given the range of different things people want out of their backups, and the bargain basement "strip out anything not absolutely required to keep the headline cost down" nature of "cloud", I would find it very surprising if there were any "cloud" providers out there doing it. – womble Jul 22 '12 at 03:30
1

Rackspace should be able to create an emergency dump of your disk image, which you can then access from your Cloud Control Panel and download to your local machine to attempt to recover data from. Someone recently asked how to do this for a Rackspace Cloud Linux image, and the answer may be helpful to you.

Community
  • 1
Michael Hampton
  • 237,123
  • 42
  • 477
  • 940
  • This seems the only helpful answer on this page and that's what I would think RS should provide us with but they keep refusing to do so. I don't understand why though, it's really hard to believe that server didn't have at least Raid5 and data is completely gone. But we not getting much explanation from Rackspace... – Elvin R. Jul 22 '12 at 20:45
  • Here's the most recent answer from Rackspace Cloud tech support (in request for the emergency dump): "Unfortunately we can not provide a dump of your disk image as there is no data we can provide. Due to an issue with the hypervisor the vhd's that contain the data for your server were deleted." – Elvin R. Jul 22 '12 at 21:17
  • Now that sucks. In that case, I think the only thing you can do is download your most recent backup image and recover what you can from that. – Michael Hampton Jul 22 '12 at 21:43