xfs
[Top] [All Lists]

Re: 2.4.16 & procps

To: Tom Schroll <ths@xxxxxxxxxxxx>
Subject: Re: 2.4.16 & procps
From: Joshua Baker-LePain <jlb17@xxxxxxxx>
Date: Tue, 11 Dec 2001 07:27:49 -0500 (EST)
Cc: Keith Owens <kaos@xxxxxxxxxxxxxxxxx>, <linux-xfs@xxxxxxxxxxx>
In-reply-to: <200112110523.AAA15349@thunder.liststorm.com>
Sender: owner-linux-xfs@xxxxxxxxxxx
On Tue, 11 Dec 2001 at 12:23am, Tom Schroll wrote

> release/update tree.  So I'll have to wait or get medieval on it with
> a pair of pliers and a blowtorch...and compile everything myself.

Wouldn't 'getting medieval' more likely involve stuff like maces, 
broadswords, and, if you're lucky, siege engines?  ;)

> and several of the patches failed.  In your opinion, is it OK to use
> the vanilla kernel instead of a RedHat tuned one?  I mean, I end up stripping
> the kernel anyways to tune it for my webserver.  Any suggestions would
> be appreciated.  Does RedHat even do that much stuff to the kernel that
> is worthwhile (I'd rather ask you guys this than the RH list -- I'll likely
> get flamed there...hehe)

Red Hat kernels are (or used to be) based on the ac tree (most of which 
has been merged into 2.4.16).  Red Hat also adds drivers not in the 
mainline kernel, and (of course) heavily QAs the kernel before release.  
The party line tends to be that if you don't need the added features, then 
there's nothing wrong with going with a mainline kernel over Red Hat's.  
If you want to stick with a heavily tested kernel, use one of the 
release kernels SGI put out -- there's a 2.4.14 kernel for the 1.0.2 
release.  I myself am (quite happily) using the vanilla 2.4.5+XFS from the 
1.0.1 release on our 560GB RAID.

-- 
Joshua Baker-LePain
Department of Biomedical Engineering
Duke University


<Prev in Thread] Current Thread [Next in Thread>