Snapshots provide means to capture and store filesystem state of an instance.
4.3.1 Types of snapshots
There are two types of snapshots used in OpenStack: image (instance) snapshots and volume snapshots. Both snapshot types can be utilized when creating a new instance, image snapshots being the more common use case for this.
Image snapshots hold the state of a given instance's root disk, and thus can typically be used as a bootable image. Image snapshots are managed under Compute | Images in the web UI. For instructions on creating image snapshots please see section Creating an image based on an existing image in the next chapter.
Volume snapshots capture the state of a persistent volume. They can be used for example for one-time backup before an upgrade operation that touches volume contents. Volume snapshots are managed under Compute | Volumes in the web UI.
4.3.2 General Considerations
At the time of writing there is one known issue related to snapshots, to be fixed in a future IaaS upgrade.
It is advisable to flush all active file system writes before taking the snapshot i.e. by running the sync command. For best results, one should stop services or suspend the instance entirely before taking the snapshot.
Please note that snapshots may not always be the optimal method of getting an instance into a predefined state. For longer term cloud usage, configuration automation tools should be considered. These mostly replace using snapshots by starting from scratch then automatically setting up relevant software and configurations. Examples of these tools are Ansible and Puppet. By using these tools one can get more of the automatization benefits of the cloud, and get the benefit of latest security updates upon instance creation (as long as CSC built images are used).
|Previous chapter||One level up||Next chapter|