Minimizing the Impact of Cloud Block Storage Maintenance

Rackspace occasionally performs maintenance to increase the stability and performance of the Cloud Block Storage platform. This article provides some tips for minimizing the impact of this maintenance on your environment.

How can you reduce the impact of the maintenance?

You can either clone and replace your Cloud Block Storage volume or stop all writes to the device until the maintenance is complete.

Clone and replace

After you clone your Cloud Block Storage volume, the resulting clone should land on an unaffected host server.

Non-system disk Cloud Block Storage volumes:

• Stop all writes to your current Cloud Block Storage volume before initiating the clone. After all writes to the current Cloud Block Storage volume stop, clone the volume. After the clone completes, unmount and detach the current Cloud Block Storage volume. Then attach the new volume (the clone).

📘

Clones are not verified. Ensure the newly-cloned volume is working before deleting your current Cloud Block Storage volume.

System disk Cloud Block Storage volumes (Boot from volume servers):

• Replace your current system disk with a clone.

📘

Clones are not verified. Ensure the newly-cloned volume is working before deleting your current Cloud Block Storage volume.

Stop all writes to the device

Non-system disk Cloud Block Storage volumes:

• Unmount and detach any Cloud Block Storage volume impacted by the maintenance. (The link includes instructions to delete Cloud Block Storage volumes—you should complete only the unmount and detach sections.)

System disk Cloud Block Storage volumes (Boot from volume servers):

• To avoid writes to the system disk, you must shut down your cloud server from the operating system (OS) at least 10 minutes before the maintenance starts. This causes your server to show as Shutoff in the MyCloud Portal. After you receive a ticket confirming that the maintenance is complete, reboot the server from the MyCloud Portal by selecting Actions> Reboot. Your server boots up and displays as Active in the MyCloud Portal.

Troubleshooting after a maintenance

Volume not appearing inside server:

• Shut the server down, detach the Cloud Block Storage volume, and attach it back to the server. Then you can start your server back up either through a reboot through the Control Panel or by issuing a start call by using the API.

Non-system disk Cloud Block Storage volumes:

• If your volume is in read-only mode, unmount the volume, run a filesystem check (Linux® with ext3/4 or fsck, Windows® with chkdsk), and then remount the volume. If the disk is still in read-only mode after that,
  then shut the server down and follow the detach and attach steps from the preceding section. If that does not work, contact Rackspace Support.

System disk Cloud Block Storage volumes (Boot from volume servers):

• Reboot the server. If the server is unresponsive or does not reboot, you can boot into rescue mode by using a specific image.

📘

This feature is not yet available from the myrackspace.com portal, so you must use the API.

When the server is in rescue mode, follow the preceding directions for non-system disk Cloud Block Storage volumes. Contact Rackspace Support for additional assistance.