scp and symlinks

[Bruce Marshall]

On Saturday 20 November 2004 10:11 am, Dr. Jones wrote:
> Jorge Almeida wrote:
> > On Sat, 20 Nov 2004, Kurt Wall wrote:
> >> A workaround is to use rsync:
> >>
> >> rsync -r -e ssh path user at host:/path
> >> rsync -r -e ssh user at host:/path path
> >>
> >> -r tells rsync to recurse into directovies. -e ssh specifies the remote
> >> shell (ssh in this case) for rsync to use.
> >
> > It works great (I added the -l flag).
> > Thank you.
>
> Just curious how rsync compares with using scp? I routinely invoke scp
> on my linux box at home to transfer files from work, instead of having
> to find the right zip disk, a la :
>
> scp scott at my.ip.numberat.work:/usr/archive/blah/blah/blah/*
> /home/scott/archive
>
> It seems to work pretty well, and I always have a backup of my data
> files. What interests me about rsync is the "sync" part of the name. I
> would like to set up a better backup and/or transfer routine that will
> just grab all the files newly created or changed on my work machine. Is
> there a way to run either rsync or scp on a batch file, to have it just
> grab the files new or altered on a given day, or even better, just the
> changed portions instead of downloading the entire files?
>
> Scott

I run rsync on a nightly basis to backup /home/  and some other partitions.

rsync -auvzr -e ssh --delete /home/   <othernamchine>:/home2/


Works fine for me but others may want to use different options on the command.  
The above command is in a crontab.