[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Savannah-hackers-public] Re: autoredirect for savannah downloads

From: Yavor Doganov
Subject: Re: [Savannah-hackers-public] Re: autoredirect for savannah downloads
Date: Sat, 01 Nov 2008 00:29:55 +0200
User-agent: Wanderlust/2.15.5 (Almost Unreal) SEMI/1.14.6 (Maruoka) FLIM/1.14.9 (Goj┼Ź) APEL/10.7 Emacs/22.3 (i486-pc-linux-gnu) MULE/5.0 (SAKAKI)

Karl Berry wrote:
> I did put a note about it onto when I added
> the anchors a couple days ago :).  

The thing is, nearly nobody uses that list (now).

> What other important places come to mind?

I vaguelly remember a discussion involving you, me and the sysadmins
to run this as an official service, and to actually
redirect to the _fastest_ nearest mirror.

> We are using it in TeX Live 

OK, that sounds compelling.  If it works for the beast, it'll work for
the little beasties too.

> What I wonder about is this scenario:

Not very relevant, since very seldom maintainers update packages on
the same day.  And even if they do, a maintainer downloading the
tarball from once or twice (or more) is perfectly fine.
What matters is users constantly doing that, and various auxilliary
distro services (such as DEP, PTS, etc.) bombarding the main server.

Of course, I don't have any data to back that up, and most definitely
I don't want to argue.  I remember a mirror admin complained that
there was zero traffic after some considerably large period, and he
had the feeling that it is useless to maintain the mirror.  If such a
facility is in place, we can make the mirrors actually useful to
users, spreading the load as in the old days.

> The multiplexer does ignore mirrors that are "too old" (more than
> two days currently)

That's the right behavior, IMHO.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]