important caveat with Rsync on NT and daylight savings time

[jw schultz]

On Tue, Oct 29, 2002 at 10:01:58PM +0100, bart.coninckx at sita.be wrote:
> 
> That's actually a very good suggestion. First I figured that in this way
> files changed within the hour after creation would be ignored, but they
> probably represent a very small minority anyway.

The number of files whose timestamp is less than one hour
since the last rsync run and still have the same size will
be very small.

> On the other hand, this does not really rectify the situation, but will
> allow us to postpone the real sync again until, let's say a holiday, where
> we have plenty of time to sync.

Reducing the window won't work (brain fart on my part).

If you run an rsync without the modify-window on
part of the trees each night you can get caught up.

> 
> On Tue, Oct 29, 2002 at 09:06:26PM +0100, bart.coninckx at sita.be wrote:
> >
> > No matter what my understanding's like: the timestamps on the files
> changed
> > after last weekend, resulting in a mismatch between source and
> destination
> > files. I've changed the clock to a timezone one hour ahead to compensate,
> > did a "F5" in a particular folder and miraculously the timestamps
> followed
> > along with one hour. Unfortunately, it was already too late for a lot of
> > files ...
> 
> In your situation you might want to use --modify-window=3600
> and each night either reduce the window or apply it to fewer
> files.  That way you can spread the load.
> 
> You'll need to do this again in April 8-(
> 
> Holloween came early for you.
> 

-- 
________________________________________________________________
	J.W. Schultz            Pegasystems Technologies
	email address:		jw at pegasys.ws

		Remember Cernan and Schmitt