[ale] How do people deal with RHEL?

Jerald Sheets questy at gmail.com
Thu Mar 24 20:22:47 EDT 2011


I was hoping someone would go in depth about it.  

This is why I'm on ALE.  I love to learn new things.


#!/jerald
Linux User #183003
Ubuntu User #32648
Public GPG Key:  http://questy.org/js.asc

-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GIT/MU d-@ s++(++)>+++:> a+ C++++(+++)$>++ UBLAVHSC++(on)$>++++ P++(+++)$>++++ L++(++++)$>+++ !E---(---)>--- W+(++)$>+++ N(+)$>++ !o !K-- w(--)>--- O()@> M++(++)$>++ V()>- PS+++()@>-- PE(++)@>+ Y+(+)@>+ PGP++(++)$>+++ t+(++)@>+++ 5(+)@>+ X+(++)@>+++ R+(+)@>++ tv-(+)$>++ b+++(++)$>++ DI++++(++)>+++ D++(++)@>++ G++(++)@>++ e++(++)$>++ h(-)$>- r+++(+++)@>+++ y+(+++)>++++@ 
------END GEEK CODE BLOCK------



On Mar 24, 2011, at 5:30 PM, Michael Trausch wrote:

> On 03/24/2011 04:36 PM, Chris Fowler wrote:
>> I think it is still considered monolithic even if they are modules and
>> linked into the kernel at run time.
> 
> Linux has historically been a monolithic kernel, but functionality has 
> been added to it that is distinctly microkernel functionality; today it 
> could be considered to be a hybrid of sorts; the style used to implement 
> kernel module loading and unloading, the FUSE and CUSE drivers, and 
> other things that pass messages back and forth between userspace and 
> kernel space can all be argued to be microkernel style functionality.
> 
> There have also been discussions in the past about modifying the 
> kernel's structure, migrating from a predominately monolithic style to a 
> predominately microkernel style.  Given that since those discussions the 
> kernel has gained quite a bit of functionality styled that way, I'd 
> wager that we'll see it gradually become more and more like a microkernel.
> 
>> A microkernel has drivers that communicate with the kernel as a form of
>> IPC.  The benefit is that failure of a driver in most cases will not
>> take down a system.  The con is that a micro with this mess of
>> communication is much harder to implement than a mono.
> 
> Depends on how you implement it.  Sometimes, as with the Hurd, the 
> hardest part is getting people to write drivers for the thing.  The 
> kernel itself is largely complete, but lacking in hardware support (and 
> thus lacking in the ability to be used on bare hardware).  However, the 
> Hurd supposedly works really well in virtual machine environments.  I 
> haven't attempted to use it in several years, however.
> 
> 	--- Mike
> _______________________________________________
> Ale mailing list
> Ale at ale.org
> http://mail.ale.org/mailman/listinfo/ale
> See JOBS, ANNOUNCE and SCHOOLS lists at
> http://mail.ale.org/mailman/listinfo




More information about the Ale mailing list