Scheduled images FAQ

Cloud Servers users can create two different kinds of images from their running servers: *manual* or *scheduled*. The user inititates manual images and runs only one time. A scheduled image, captured automatically either daily or weekly, retains up to the number of images specified by the user.

By capturing images automatically versus manually triggering them each time, you create an image history. You can use that history to recover your server to the point of the image in the case of emergency or server failure.

In the following cases, scheduled images might not be the best choice:

• Images are useful for recovery in many scenarios, but you should never use them as your sole source of recovery. We recommend that you use Cloud Backup or your preferred backup method in conjunction with server images.
• Some application servers are not good candidates for scheduled images. In particular, some database management systems need to be in a particular state before taking an image if you want a working DBMS when you boot a server from the image. Consult your DBMS vendor for more information. If you have such a DBMS and want to create an image of the server, read Using task states with server imaging.
• Both manual and scheduled images are images of the system disk only.

Initiating scheduled images is easy and available to both Cloud Control Panel and API users:

• Cloud Control Panel: Click the cog (gear wheel) next to your server name, and select Schedule Image. From the drop-down menu, select whether you want to create an image every Day or Week. If you select Week, you can also specify on which Day of Week for the image creation. Finally, specify the number of images you want to retain and then click Create Schedule. To learn more about the Cloud Control Panel, see Introducing the Rackspace Cloud Control Panel.
• API users: See the Cloud Servers API documentation for the Scheduled Images extension.

We charge you for images, stored in **Cloud Images**, as if we stored them in your Cloud Files account. Thus, we charge your normal Cloud Files rate, including any applicable tiering or volume discounts.

The system creates scheduled images with a name based on the following scheme:

• daily-{server-name}-{10-digit-number} (for daily scheduled images)

• weekly-{server-name}-{10-digit-number} (for weekly scheduled images)

  Note: An image name is limited to 255 characters, so if you have a server name longer than 238 characters, the system truncates it to fit.

The best way to tell is to look at the image metadata, but you can't see it in the Cloud Control Panel. By using the API, you can see that an image created by the scheduled images service has the following metadatum: org.openstack__1__created_by: scheduled_images_service

No (see next answer), but daily scheduled images are roughly 24 hours apart and occur on different dates UTC.

For weekly images, you specify the day of the week (determined by UTC) when you'd like your server image created. Because some days of the week are much more popular than others for scheduling images, in rare circumstances we might create your weekly scheduled image in a window beginning 12:00 UTC the day before the day of the week you specify and ending at 12:00 UTC the day after the day of the week you specify.

To operate the scheduled images service with minimal impact on on-demand snapshots and other network-intensive data transfers, we might change the time a snapshot occurs at any time to optimize load. We reserve the right to modify the scheduled time of your image to balance the number of image creations in flight across the cloud and throughout the day.

The implementation uses the normal OpenStack snapshot process, so the created\_at timestamp is when the call occurred at the start of the snapshot. The availability time of the image (when its status is ACTIVE) depends on factors such as the size of the image and overall network congestion in the cloud.

You can't tell the service to skip a day. We do not recommend trying to do this manually by disabling scheduled images on a server and then re-enabling them because of the dynamic nature of the scheduling system. It is better just to delete any scheduled images you don't want in your account.

However, the service might miss capturing a scheduled image on a particular day for a server. This could happen if your server is in a state that does not allow the system to make snapshots (for example, if you are doing a server resize at that time).

The service attempts to create a snapshot of your server three times, with the re-tries approximately one hour apart. (**Note**: We are constantly monitoring the service faults and might change the number or frequency depending on how the service is performing.)

If you notice two or more consecutively skipped days, contact Rackspace support.

The retention value is the maximum number of scheduled images for that particular server that your account retains. After the number of scheduled images for this server reaches the retention value, the system deletes the oldest scheduled images so that the total number of scheduled images for this server does not exceed the retention value.

Because you set the retention value on a per-server basis, you can specify a different retention value for each server.

The maximum retention value in the Rackspace Open Cloud is 65535. That's about 179 years of daily images.

When a scheduled image of a server has successfully completed, the scheduled images service creates a list of all the scheduled images for that server and, if necessary, deletes the oldest images until it reaches the retention value for that server.

Let's suppose that we're talking about the scheduled images for server `d615a437-aaa9-4a52-a1c0-5bcb0d33038c`. To determine how many scheduled images are in your account for this server, the scheduled images service looks only at images that meet the following two constraints:

• The image has the metadatum org.openstack__1__created_by with the value scheduled_images_service.
• The image property instance_uuid has the value d615a437-aaa9-4a52-a1c0-5bcb0d33038c.

So, if you remove the metadatum org.openstack__1__created_by from the scheduled image that you want to save, the system does not perform retention culling on the snapshot.

You cannot modify image metadata by using the Cloud Control Panel. Instead, you can use API calls.

You can change the retention value by using the following tools:

• Cloud Control Panel
• API

Just specify the new value you want to use for the retention.

Nothing at first, but when your next scheduled image finishes, the scheduled images service uses the new retention value when it calculates whether to delete any images from your account.

A scheduled image is just a normal snapshot. You can do what you like with it just as you can with your other snapshots. You can delete a scheduled image at any time by using your normal workflow, such as the Cloud Control Panel, direct API calls, and so on.

You cannot specify a particular time for your server snapshot, and we can't guarantee what time your scheduled image becomes active. We've placed these restrictions because we spread out the scheduled times for snapshots to avoid interfering with each other or with on-demand snapshots. The time before an image becomes active depends on the current network traffic load, among other things. We guarantee that all users receive the same best-effort service.

In using scheduled images, keep the following in mind:

• Smaller snapshots tend to finish more quickly.
• A very large snapshot might take so long to finish that it could block the next day's scheduled snapshot from occurring.
• If you have a large amount of data to save, you can explore other backup options

There are no plans to change the Classic service.

For more information, see the Scheduled Images API Extension Documentation.