[Gluster-users] Architecture advice

[Gandalf Corvotempesta]

Let's assume that we have two geographically distand datacenter.
I have to create a redundant gluster infrastructure for a cluster of webservers.
Sites will be accessibled from both location at the same time.

With georeplication i'll be able to sync data from the master site to
the slave site.

But:
1) are data replicated with georeplication readable on the slave site?
2) how can I manage writes coming from the second site? AFAIK the
georeplication is only in one direction "Master" -> "Slave"

Bot sites should be write-active at the same time.
Any advice?