Storage architecture ==================== OpenStack has multiple storage realms to consider: * Block Storage (cinder) * Object Storage (swift) * Image storage (glance) * Ephemeral storage (nova) Block Storage (cinder) ~~~~~~~~~~~~~~~~~~~~~~ The Block Storage (cinder) service manages volumes on storage devices in an environment. In a production environment, the device presents storage via a storage protocol (for example, NFS, iSCSI, or Ceph RBD) to a storage network (``br-storage``) and a storage management API to the management network (``br-mgmt``). Instances are connected to the volumes via the storage network by the hypervisor on the Compute host. The following diagram illustrates how Block Storage is connected to instances. .. figure:: ../figures/production-storage-cinder.png :width: 600px The diagram shows the following steps. +----+---------------------------------------------------------------------+ | 1. | A volume is created by the assigned ``cinder-volume`` service | | | using the appropriate `cinder driver`_. The volume is created by | | | using an API that is presented to the management network. | +----+---------------------------------------------------------------------+ | 2. | After the volume is created, the ``nova-compute`` service connects | | | the Compute host hypervisor to the volume via the storage network. | +----+---------------------------------------------------------------------+ | 3. | After the hypervisor is connected to the volume, it presents the | | | volume as a local hardware device to the instance. | +----+---------------------------------------------------------------------+ .. important:: The `LVMVolumeDriver`_ is designed as a reference driver implementation, which we do not recommend for production usage. The LVM storage back-end is a single-server solution that provides no high-availability options. If the server becomes unavailable, then all volumes managed by the ``cinder-volume`` service running on that server become unavailable. Upgrading the operating system packages (for example, kernel or iSCSI) on the server causes storage connectivity outages because the iSCSI service (or the host) restarts. Because of a `limitation with container iSCSI connectivity`_, you must deploy the ``cinder-volume`` service directly on a physical host (not into a container) when using storage back ends that connect via iSCSI. This includes the `LVMVolumeDriver`_ and many of the drivers for commercial storage devices. .. note:: The ``cinder-volume`` service does not run in a highly available configuration. When the ``cinder-volume`` service is configured to manage volumes on the same back end from multiple hosts or containers, one service is scheduled to manage the life cycle of the volume until an alternative service is assigned to do so. This assignment can be made through the `cinder-manage CLI tool`_. This configuration might change if `cinder volume active-active support spec`_ is implemented. .. _cinder driver: http://docs.openstack.org/developer/cinder/drivers.html .. _LVMVolumeDriver: http://docs.openstack.org/developer/cinder/drivers.html#lvmvolumedriver .. _limitation with container iSCSI connectivity: https://bugs.launchpad.net/ubuntu/+source/lxc/+bug/1226855 .. _cinder-manage CLI tool: http://docs.openstack.org/developer/cinder/man/cinder-manage.html#cinder-volume .. _cinder volume active-active support spec: https://specs.openstack.org/openstack/cinder-specs/specs/mitaka/cinder-volume-active-active-support.html Object Storage (swift) ~~~~~~~~~~~~~~~~~~~~~~ The Object Storage (swift) service implements a highly available, distributed, eventually consistent object/blob store that is accessible via HTTP/HTTPS. The following diagram illustrates how data is accessed and replicated. .. figure:: ../figures/production-storage-swift.png :width: 600px The ``swift-proxy`` service is accessed by clients via the load balancer on the management network (``br-mgmt``). The ``swift-proxy`` service communicates with the Account, Container, and Object services on the Object Storage hosts via the storage network(``br-storage``). Replication between the Object Storage hosts is done via the replication network (``br-repl``). Image storage (glance) ~~~~~~~~~~~~~~~~~~~~~~ The Image service (glance) can be configured to store images on a variety of storage back ends supported by the `glance_store drivers`_. .. important:: When the File System store is used, the Image service has no mechanism of its own to replicate the image between Image service hosts. We recommend using a shared storage back end (via a file system mount) to ensure that all ``glance-api`` services have access to all images. Doing so prevents losing access to images when an infrastructure (control plane) host is lost. The following diagram illustrates the interactions between the Image service, the storage device, and the ``nova-compute`` service when an instance is created. .. figure:: ../figures/production-storage-glance.png :width: 600px The diagram shows the following steps. +----+---------------------------------------------------------------------+ | 1 | When a client requests an image, the ``glance-api`` service | | | accesses the appropriate store on the storage device over the | | | storage network (``br-storage``) and pulls it into its cache. When | | | the same image is requested again, it is given to the client | | | directly from the cache. | +----+---------------------------------------------------------------------+ | 2 | When an instance is scheduled for creation on a Compute host, the | | | ``nova-compute`` service requests the image from the ``glance-api`` | | | service over the management network (``br-mgmt``). | +----+---------------------------------------------------------------------+ | 3 | After the image is retrieved, the ``nova-compute`` service stores | | | the image in its own image cache. When another instance is created | | | with the same image, the image is retrieved from the local base | | | image cache. | +----+---------------------------------------------------------------------+ .. _glance_store drivers: http://docs.openstack.org/developer/glance_store/drivers/ Ephemeral storage (nova) ~~~~~~~~~~~~~~~~~~~~~~~~ When the flavors in the Compute service are configured to provide instances with root or ephemeral disks, the ``nova-compute`` service manages these allocations using its ephemeral disk storage location. In many environments, the ephemeral disks are stored on the Compute host's local disks, but for production environments we recommend that the Compute hosts be configured to use a shared storage subsystem instead. A shared storage subsystem allows quick, live instance migration between Compute hosts, which is useful when the administrator needs to perform maintenance on the Compute host and wants to evacuate it. Using a shared storage subsystem also allows the recovery of instances when a Compute host goes offline. The administrator is able to evacuate the instance to another Compute host and boot it up again. The following diagram illustrates the interactions between the storage device, the Compute host, the hypervisor, and the instance. .. figure:: ../figures/production-storage-nova.png :width: 600px The diagram shows the following steps. +----+---------------------------------------------------------------------+ | 1 | The Compute host is configured with access to the storage device. | | | The Compute host accesses the storage space via the storage network | | | (``br-storage``) by using a storage protocol (for example, NFS, | | | iSCSI, or Ceph RBD). | +----+---------------------------------------------------------------------+ | 2 | The ``nova-compute`` service configures the hypervisor to present | | | the allocated instance disk as a device to the instance. | +----+---------------------------------------------------------------------+ | 3 | The hypervisor presents the disk as a device to the instance. | +----+---------------------------------------------------------------------+