Re: comparison of star and rsync methods
Les Mikesell <[email protected]> Fri, 14 Mar 2008 13:02:33 -0500
| Newsgroups | gmane.comp.archivers.star.user |
|---|---|
| Message-ID | <[email protected]> |
Lasse Kliemann wrote:
>
> So the real competitors are 'rsync S --> V' and 'star S' (hoping that the
> problems with the latter can be resolved). Clearly, star is faster than
> rsync; the star runs only take about 60% of the time of the rsync runs.
I think your results with rsync will vary with the number of files in
the run, and with -H also with the number of hard links. Rsync will
read the entire directory contents with a certain amount of RAM overhead
per file before starting. If you have a few big files, that's not a
problem - with a lot of tiny files it can be noticeable. I think the
very latest version may start the comparison before the full directory
load completes, at least when running over a network, though.
> The
> big structural difference is that with rsync, one has the backup in a
> filesystem instead of in a single file (or in few files with incremental
> dumps). This can be an advantage if one often has to access particular files
> in the backup. It can also be a disadvantage if one wishes to write the
> backup to some external medium or push it over the network.
Actually the way rsync works can be a big advantage when writing over a
network that is slow compared to the disks. You can use the -z option
for compression, and it will walk through the file doing block checksum
compares with the existing remote copy, sending only the changed parts
for the remote to assemble.
--
Les Mikesell
[email protected]