[Gluster-users] replication problem

[Anand Avati]

> If i'm not wrong it will be synced once you try to access any
> of the unsynced files, except if the file does not exist on the
> first volume

That is not quite true, if all the modifications were performed from
the mountpoint, then it does not matter which server went down. If
files were removed in the backend of the first server directly (not
via glusterfs mountpoint), then it does not appear in the 'ls' output.
However accessing the file by name should fix it anyways.

Avati