On Mon, Feb 27, 2006 at 05:05:39PM +0100, Alexander Skwar wrote
> El Nino wrote:
>
> > is there a way to accelerate the fetching part of emerge
> > by using prozilla or some other tool?
>
> I never quite understood the sense in those tools.
>
> Why should "prozilla or some other tool" make the
> download be faster? When I download something with
> wget, or watch emerge invoking wget, it's always
> maxing out the saturation of the line.
I have a different interpretation. Assume I'm doing an
emerge --deep --update --world
I've set my ADSL router-modem to log off after 15 minutes. The
sequence of events is something like...
1) emerge wants to download a package, so it attempts to connect to a
server, taking several seconds to wake up my ADSL router-modem
2) emerge downloads a big package, taking a few minutes to do so
3) emerge spends the next half hour building the big package, during
which time the modem-router logs off
4) emerge finally finishes building the package, and wants to work on
the next one... GOTO 1
I could run a short script
#!/bin/bash
emerge --deep --update --world --fetchonly
emerge --deep --update --world
...but I'd like to get an emerge going on the 1st package as soon as
it's finished downloading, whilst having the downloads of all the other
packages continue in a separate thread. When the 1st build is finished,
check whether the 2nd package has been downloaded. If not, wait. Then
build the 2nd package... etc, etc.
The best way to describe it is as a --fetchonly emerge that launches a
separate emerge as each individual package is finished downloading. The
"build" emerges should be serialized, i.e. only one build running at a
time, because a package may depend on the immediately preceding package.
--
Walter Dnes <[EMAIL PROTECTED]> In linux /sbin/init is Job #1
My musings on technology and security at http://tech_sec.blog.ca
--
gentoo-user@gentoo.org mailing list
