On Tuesday 31 March 2009 21:39:59 G. Matthew Rice wrote:
> I'm not certain that I agree 100% with no hands-on compilation knowledge
> but I also have to admit that I haven't had to compile a kernel in a number
> of years.  Some people do compile them if they need the very latest
> releases for some reason.  I sure don't volunteer for that work, though.
>
> That said, I'm open to the idea of dropping more Kernel building objectives
> from the exam but that will have to wait for the objective refresh in
> 18-24-ish months from now.

I'd argue that at LPIC-2 level it's an important skill to be able to patch 
specific drivers, or add an out-of-tree driver, and compile the resulting 
kernel. OTOH, selecting obscure kernel options to tweak the last 0.001% of 
performance is rightly the province of the distro kernel maintainer and isn't 
something we should be *expecting* a sysadmin to know how to do well.

Case in point: I have boxes that have disgusting performance using the distro-
stock driver to connect to the SAN. There is no officially maintained driver 
available but a quick recompile gets me the testing version which works well. 
Admittedly this is a FreeBSD box but the principle holds - someone in my team 
has had to do this on two occasions in the last year that I know of on Linux.

And as for desktop support - on new notebooks it's almost a prerequisite to 
get anything brand new really working. iwl3945, thinkfinger, cisco VPN, TV 
cards - I've had to do them all on off-the-shelf Dell notebooks.

To sum up: I definitely expect an LPIC-2 qualified person to be able to find 
trustworthy code via Google, download it and run make to get needed hardware 
working (Only Buy Supported Hardware! doesn't always work out - purchasing 
personnel don't always have the same point of view as us).

-- 
alan dot mckinnon at gmail dot com
_______________________________________________
lpi-examdev mailing list
[email protected]
http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev

Reply via email to