Basic concepts of CEPH storage


I. distributed file system:

It means that the physical storage resources managed by the file system are not necessarily connected to the local node directly, but to the node through the computer network.

Design basis and C / S architecture of distributed file system (client / server)

Common distributed file systems: CEPH, (red hat) Hadoop, fastdfs (domestic)

CEPH distributed file system

Features: high expansion, high availability, high performance.

It can provide block storage, file system storage and object storage.

It can provide Pb level storage space (1PB = 1024tb = 1024 * 1024gb)

At present, software definition storage as a major development trend of storage industry, more and more recognized by the market.

(software defined storage: it is a storage method of data. A set of software written in C language or other programming languages is used to integrate scattered hard disks or disk arrays to improve storage services)

Ceph component

OSDs: storage devices monitors: cluster monitoring components radosgateway (rgw): object storage gateway

MDSS: store metadata of file system (this component is not required for object storage and block storage)

Client: CEPH client

3. CEPH block storage (Rados):

Integrated in the kernel, it provides snapshot clone and other functions, supports memory cache, which greatly improves the performance.

Single block device: CD, disk

Distributed block devices: CEPH, cinder

IV. CEPH file system storage (cephfs)

Metadata (metadata)

The data in the file system is divided into data and metadata.

Data refers to the actual data in ordinary files

Metadata refers to the system data used to describe the characteristics of a file.

Such as: access rights, file owners, etc.

V. object storage

All data in the object store is considered as an object, so any data can be stored in the object store server. (API provider required)

Renew after detailed configuration