Hi all,

I've been Googling this problem for a bit and can't find any existing
standard solutions, so I was hoping one of the gurus out there might
have some suggestions.

I want to have a ZFS filesystem replicated between 3 sites in a
master-master-master configuration: whenever a file gets
created/deleted/modified on any one of the servers this change gets
pushed to the other two. Each site should be able to carry on operating
(albeit out-of-sync) in the event of an outage to either or both of the
others, queuing the outstanding modifications until the outage is fixed.
This clearly removes the possibility of having a central server which
coordinates updates.

Is there any known way of doing this on Solaris without hacking the
filesystem driver? The filesystems in question are pretty large (>2Tb),
so a cronned rsync is out of the question. Most of the solutions out
there seem to be for active-passive or master-slave replication, which
is of no use in this situation.

Any thoughts or suggestions gratefully received.

Chris

_______________________________________________
opensolaris-discuss mailing list
[email protected]

Reply via email to