Chapter 4. Configuring DRBD

The procedures in this chapter configure LVM and DRBD for a two-node high-availability cluster.

The following list identifies the assumptions that inform the DRBD resource definition for Control Center:

  • Each node has either one or two NICs. In dual-NIC hosts the private IP/hostnames are reserved for DRBD traffic. This is recommended configuration, which enables real-time writes for disk synchronization between the active and passive nodes, and no contention with application traffic. However, it is possible to use DRBD with a single NIC.
  • All volumes should synchronize and failover together. This is accomplished by creating a single resource definition.
  • DRBD stores its metadata on each volume (meta-disk/internal), so the total amount of space reported on the logical device /dev/drbdn is always less than the amount of physical space available on the underlying primary partition.
  • The syncer/rate key controls the rate, in bytes per second, at which DRBD synchronizes disks. Set the rate to 30% of the available replication bandwidth, which is the slowest of either the I/O subsystem or the network interface. The following example assumes 100MB/s available for total replication bandwidth (0.30 * 100MB/s = 30MB/s).