Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.
> Am 04.04.2017 um 14:12 schrieb Igor Cicimov <igorc at encompasscorporation.com>: > > Or simply use different host names for the other network like: > > 192.168.1.1 fs1 > 192.168.1.2 fs2 > 192.168.1.3 fs3 > 10.10.10.1 sfs1 > 10.10.10.2 sfs2 > 10.10.10.3 sfs3 > > and set the cluster using those names: > > drbdmanage init 10.10.10.1 > drbdmanage add-node sfs2 10.10.10.2 > drbdmanage add-node sfs3 10.10.10.3 > > That's what I usually do it also helps to differentiate which network are we talking about from the host name perspective. Except if keeping the original names is a must for some reason ... That’s what I tried, but what is not working, because the drbdmanage software detects its own name by doing os.uname(). And that reports the name from /etc/hostname, corresponding to the external interface. It gives me a half working cluster where I can see the drbd resources correctly but the quorum mechanism fails. (Each node seeing the others offline)