Hi,
We use drbd version 9.0.21-1 on CentOS 8 machine (kernel version
4.18.0-147.5.1.el8_1.x86_64) in two nodes. The direct link between the
two nodes is ensured by a fiber connection.
Some times, the direct link is down and the cluster switch drbd disks
from one node to the other and get at the end StandAlone state (split
brain).
DRBD disk synchronization is ensured *only* by the direct link based on
the configuration below :
/resource test {//
// device minor 1;//
// net {//
// shared-secret xxxxxxxxxxxxxxxx;//
// }//
// on node1 {//
// address 192.168.1.1:7771;//
// disk /dev/vg0-node1/test;//
// flexible-meta-disk /dev/vg0-node1/test-md;//
// }//
// on node2 {//
// address 192.168.1.2:7771;//
// disk /dev/vg0-node2/test;//
// flexible-meta-disk /dev/vg0-node2/test-md;//
// }//
// }/
Does DRBD support synchronization over more than one network in order to
keep synchronization even the direct link (network 192.168.1.0/30 in our
case) is down ? How can we define two addresses in each node ?
walim
_______________________________________________
Star us on GITHUB: https://github.com/LINBIT
drbd-user mailing list
[email protected]
https://lists.linbit.com/mailman/listinfo/drbd-user
