Comments on: Optimizing Gentoo (Part 2) https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/ Daily tips, tricks, and news about Linux and Linux-related software. Sun, 22 Oct 2006 23:32:57 +0000 http://wordpress.org/?v=1.5.1.3 by: Jesper Frickmann https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-2323 Tue, 05 Apr 2005 12:39:48 +0000 https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-2323 That reduced my boot time from 1:14 to 1:10. Great Gentoo for "Ricers" ;-) That reduced my boot time from 1:14 to 1:10. Great Gentoo for “Ricers” ;-)

]]>
by: Paul Pacheco https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-2298 Mon, 04 Apr 2005 13:12:46 +0000 https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-2298 <!-- spamk : Waaaay too many links: 8 --> <!-- spamk : Comment on old post. --> <!-- spamk : CAPTCHA: approved --> <!-- spamk : KARMA: -0.5 --> <!-- spamk : Treatment: captcha --> I have worked very hard on this area and your article has a number of problems and could go even further. <a href="http://bugs.gentoo.org/show_bug.cgi?id=69579">here</a> I keep track of all this type of things. In particular: 1) mounting in parallel does not work if you have say /, /usr and /usr/local in different devices, you may end up mounting /usr/local before /usr, man mount has more details on this. So you should at least add a warning. 2) RC_PARALLEL_STARTUP="yes” is not very good: at any given point it is doing 2 tasks or 1 task and a busy loop, so it is not very parallel. I have written an appropiate patch that is trully parallel with no busy wait. You can find it <a href="http://bugs.gentoo.org/show_bug.cgi?id=69854">here</a> 3) The env-update thing <a href="http://bugs.gentoo.org/show_bug.cgi?id=71285">has been fixed</a> some time ago. So your change does not do anything. 4) Your modules-update patch has number of issues and is just broken. Check <a href="http://bugs.gentoo.org/show_bug.cgi?id=55329"> Here </a> for a discution about them and the proper fix. 5) If you applied my parallel patch, try to remove unneded stuff from boot runlevel, such as alsa. 6) Remove consolefont, it is not needed on x86 systems. 7) I have also written a <a href="http://bugs.gentoo.org/show_bug.cgi?id=70689">patch to start xdm before other services</a>. Just adding it to boot won't do it. This is specially good after my parallel startup patch. 8) You may want to make a chart out of your boot process to see what the bottleneck is. I made an <a href="http://bugs.gentoo.org/show_bug.cgi?id=74425">ebuild</a> for that. Check <a href="http://bootchart.sourceforge.net/samples.html">Here</a> for some samples. This will conflict with the parallel startup patch. If you want both, just mail me. All of this has been discussed to death in <a href="http://forums.gentoo.org/viewtopic-t-131142-highlight-.html">Here</a>. Search for the posts from "castorilo" (myself) in page 5 for nice instrucctions.


I have worked very hard on this area and your article has a number of problems and could go even further.

here I keep track of all this type of things.

In particular:

1) mounting in parallel does not work if you have say /, /usr and /usr/local in different devices, you may end up mounting /usr/local before /usr, man mount has more details on this. So you should at least add a warning.

2) RC_PARALLEL_STARTUP=”yes” is not very good: at any given point it is doing 2 tasks or 1 task and a busy loop, so it is not very parallel. I have written an appropiate patch that is trully parallel with no busy wait. You can find it here

3) The env-update thing has been fixed some time ago. So your change does not do anything.

4) Your modules-update patch has number of issues and is just broken. Check Here for a discution about them and the proper fix.

5) If you applied my parallel patch, try to remove unneded stuff from boot runlevel, such as alsa.

6) Remove consolefont, it is not needed on x86 systems.

7) I have also written a patch to start xdm before other services. Just adding it to boot won’t do it. This is specially good after my parallel startup patch.

8) You may want to make a chart out of your boot process to see what the bottleneck is. I made an ebuild for that. Check Here for some samples. This will conflict with the parallel startup patch. If you want both, just mail me.

All of this has been discussed to death in Here. Search for the posts from “castorilo” (myself) in page 5 for nice instrucctions.

]]>
by: Alexander Davidsen https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-641 Fri, 04 Mar 2005 08:59:54 +0000 https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-641 This was a really nice articel. Alot of usefull tips. Have you posted this in the gentoo forums, if not, do it :P This was a really nice articel. Alot of usefull tips. Have you posted this in the gentoo forums, if not, do it :P

]]>
by: je https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-637 Fri, 04 Mar 2005 01:05:13 +0000 https://linux-blogger.com/2005/02/22/optimizing-gentoo-part-2/#comment-637 NIce tips NIce tips

]]>