Mag Gam wrote: > Hi All, > > I am trying to understand the purpose of rsyncd. Why does it exist? > Why not just use rsync ad-hoc or via script? > well, if you use rsync -> rsyncd, there's no encryption, so less CPU overhead, so faster transfers, but there's also less security, so this should only be used on a secure local network. the normal way most folks use rsync is over ssh, where the sshd at the remote end invokes another copy of rsync for you to recieve the commands. > Also, to a more serious note. I need to keep a filesystem on 1 server > and another server synced by. > What is the best way to do this? I want to have /source and /target to > be exactly the same. I don't want any extra files in /target. I want > the timestamps to be the exact same in /target. Does anyone have a > good rsync arguments I can use? I have been using -avzl but I was > wondering if there was anything better. > you might look into drbd, although the target file system shouldn't be mounted while drbd is active. if you need to read /target at the same time, forget it, stick with batch rsync at some interval. drbd does online block level replication in near realtime, so if the source server crashes, the target usually has everything within a few milliseconds. Its most frequently used in conjunction with heartbeat to implement a HA solution without expensive sharable disk storage.