Re: ftp.kernel.org overloaded :-(
Jeremy Katz (katzj@technojunkie.com)
Sat, 29 Aug 1998 09:37:54 -0700 (PDT)
On Fri, 28 Aug 1998, Joel Jaeggli wrote:
> the mirror at:
>
> ftp://limestone.uoregon.edu/pub/linux/kernel/
>
> seems to have picked up everything up to pre-patch-2.1.120-1 as of 6pm
> august 28... It runs on a 12 hour cycle, and the way I have it set up at
> the moment it seems to be keeping up with the rapid changes relativly
> well...
>
> On Fri, 28 Aug 1998, Gerd Knorr wrote:
>
> >
> > Hi !
> >
> > I have some problems to pick up the 2.1.118 patch. Downloading directly
> > from frp.kernel.org is out of question becauce it takes way to long.
> > Seems even the mirrors have problems to pick up the patches. My favorite
> > Mirrors (ftp.cs.tu-berlin.de,ftp.gwdg.de) still hav'nt 2.1.118. 2.1.119
> > is already there, probably becauce it is smaller and therfore easier to
> > download.
> >
> > Can this be fixed somehow? Maybe limit access to ftp.kernel.org for
> > "normal" users and have extra ftp class for mirrors?
>
> HPA discussed this at some point. I haven't heard anything back on the
> issue since... part of the problem being that a full mirror is currently
> running around 4GB and seems to be growing by 10 or 20MB every other day
> at this point.
>
> [joelja@limestone kernel]$ du
[snip of the really long du]
> 3800022 .
>
Okay, how's this for a thought... Linus has said many times that he would
prefer to distribute _just_ patches. This being the case, what if release
of devel kernels was changed so that every ten or so kernels, a complete
tarball would be released and just release the patches in between (though
have the patches available for the decade patches). This would cut down
on the amount of space required for a complete kernel.org mirror I would
think. Another thought would be to set up a machine patches.kernel.org so
that people downloading the smaller file can get it and go without having
to compete with people downloading the entire tarball. Perhaps Redhat or
someone would be willing to set up such a machine. Maybe I'm just
rambling, but I've had this disucssion about kernel.org's speed with a few
people and this is what we thought might work.
Jeremy Katz
katzj@shells.technojunkie.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html