1. DRBD distributed storage solution
1. Introduction to DRBD
The full name of DRBD is: Distributed Replicated Block Device (DRBD) distributed block device replication. DRBD is composed of kernel modules and related scripts to build high-availability (HA) clusters. Its implementation is to mirror the entire device through the network, which allows users to build a real-time mirror of the local block device on a remote machine. Combined with a heartbeat connection, it can also be thought of as a type of network RAID.
DRBD has two concepts, DRBD replication (data data is replicated to different nodes), and DRBD synchronization (synchronization to prevent data inconsistency).
DRBD is responsible for receiving data, writing the data to local disk, and sending it to another host. Another host then saves the data to its own disk. Currently, DRBD only allows read and write access to one node at a time, which is sufficient for a typical failover high-availability cluster. Future releases will support two nodes for read and write access.
A DRBD system consists of more than two nodes. Similar to an HA cluster, there are also active nodes and standby nodes. On the node with the main device, applications and operating systems can run and access the DRBD device.
The data written in the master node is stored in the disk device of the master node through the DRBD device. At the same time, this data will be automatically sent to the corresponding DRBD device of the backup node, and finally written into the disk device of the backup node. On the backup node, DRBD just writes data from the drbd device to the disk device of the standby node.
Most existing high-availability clusters use shared storage, and DRBD can also be used as a shared storage device. Using DRBD does not require any hardware investment. Because it runs in IP network (IP SAN). Therefore, using DRBD as a shared storage device will save a lot of cost, because the IP network is more economical than a dedicated storage network in terms of price. ,
exist