Home > News > Content

Save The Storage Structure Of The Storage Containers

Jun 15, 2017

   Containers in recent years, the rise of application development, delivery and operation and maintenance of the way. With the application of the model in depth, has been from the original state of the main application, and gradually developed to the state of the container application. Stateful container applications must have the ability to persist data, the storage model as shown below.Storage Containers The application that runs in the container, which actually needs to be saved, can be written to a persistent Volume volume. Since containers with micro service architecture are mostly distributed systems, containers can dynamically start, stop, stretch, or migrate in multiple nodes, so when the container has persistent data, it must be ensured that the data can be Different nodes are accessed. On the other hand, the container is the application-oriented runtime environment, and the data is usually saved to the file system, where the storage interface is more suitable for application access. Combined with the above requirements, the container platform should be appropriate to have a shared file interface storage system. Container storage control plane, the control plane is usually implemented by the software, mainly to receive the North to the API received storage requests, such as create, delete, load, uninstall and migrate data volumes, and passed to the underlying data plane to complete the actual storage operation. Control plane generally need to meet the call layer specification API, either Docker container engine volume plug-in API, it can be Kubernetes and other scheduling scheduling platform Volume API.Storage Containers  Storing the data plane provides the actual persistence of the data, not only the need to achieve the container data volume read and write, snapshot, data protection and replication and other storage operations, and usually need to have multi-node data sharing capabilities. Data plane can be based on file systems such as NFS, CephFS, etc. to do the actual data storage, but also based on block devices such as iSCSI, etc. with the file interface to achieve data storage. From the deployment architecture point of view, container storage can be divided into centralized and distributed two. Centralized storage, including traditional SAN as the representative of the block storage and NAS as the representative of the file storage. Distributed storage mainly includes shared file systems such as GlusterFS and Ceph RBD block devices, in the cloud environment also includes a variety of cloud storage, such as AWS EBS block storage or S3 object storage.Storage Containers