Re: [gentoo-dev] Proposal for an alternative portage tree sync method

[Nick Rout]

It sounds interesting Ricardo. May I suggest that if you have the
resources you set it up on your own LAN or something, then compare the
results with rsync. 

You may need to get a few friends to do it over the net too, in order to try it on something less traffic friendly than your LAN.


On Tue, 22 Mar 2005 23:58:19 +0000
Ricardo Correia wrote:

> On Tuesday 22 March 2005 13:55, Patrick Lauer wrote:
> >
> > A few problems:
> > - that .iso and the .zsync metadata need to be generated. More load on
> > master server
> > - isos don't allow easy access, e.g. writing a few bytes for a tricial
> > bugfix
> > - mkisofs might shuffle the data so that transferring one large file
> > might cause more traffic than rsync does now
> >
> > I don't see the advantages over tar + binary diffs.
> >
>
> You make valid points, but notice:
> - The .zsync metadata doesn't have to be generated on the master server.
> Anyone can do it right now.
> - ISO's would have to be regenerated periodically. This could vary from every
> 30 minutes to only once per day, we'd have to see how it works.
> Personally I think every 30 minutes would be viable, but it's not really
> necessary. Once per day would be enough and better than emerge-webrsync..
>
> The advantage over tar + binary diffs:
> - Client doesn't have to remove entire portage tree and extract the tar file
> every sync.
> - I think xdelta might be possible, but bsdiff would be impossible due to the
> memory requirements for a tar this large. I don't really know how xdelta
> performs CPU-wise and memory-wise..
> - It's simpler (only 2 files on the server and very few commands
> necessary) :-)
> --
> gentoo-dev@g.o mailing list

-- 
Nick Rout
Barrister & Solicitor
Christchurch
<http://www.rout.co.nz>
<nick@×××××××.nz>

--
gentoo-dev@g.o mailing list