[lpi-discuss] Re: LPIC-3 core exam -- I am an idiot

Bryan J. Smith b.j.smith at ieee.org
Tue Dec 5 17:37:32 EST 2006

One the smartest and most experienced members in my local Orlando LUG
always starts his e-mails best with ...
  "I am an idiot ..."
Or similar variations.

To that end, "I am an idiot" because I don't agree with the following
from Etienne and Alan ... but thanx for "setting me straight."

Etienne Goyer <etienne.goyer at outlands.ca> wrote:
> Most "enterprise" (that term should really be defined
> unambiguously) already have large Active Directory or Netware
> deployment, and mostly do not care about Samba for file/print
> service (which is the topic of exam 302).

Corporations do not deploy Netscape Directory Server and Certificate
Services, or any infrastructure capability, such as Sun One, that
uses Netscape Directory Server.  And the most definitely do not
deploy OpenLDAP, which shares the same base, WU origins from well
before ADS.

Corporations do not designed a open, proper, tiered DNS tree with
Kerberos realms and other naming, object and authentication
approaches, using anything but proprietary or orphan/hostageware
infrastructure solutions.  Open systems-based DNS is very limited to
those capabilities outside of LAN/WAN, more DMZ/Internet focused.

Corporations mainly deploy proprietary eDirectory or
orphan/hostageware ActiveDirectory.  They only use Novell and Windows
for file services.  Open solutions, such as Linux services, are only
border solutions and glue options -- or those services dedicated to
database or other, strictly separate applications.

So there little justification that anyone at the LPIC-3 level should
be tested on an extensive, flexible, open network infrastructure --
including network authentication, directory (outside of basic
LDAP/LDIF already in LPIC-1/2), naming (outside of limited DNS which
is already in LPIC-1/2), objects and definitely not network file
services (also introduced in LPIC-1/2).

Corporations just don't deploy an open network stack of such
services.  They rely on vendor products and their partner products
and services for capabilities and integration.  We shouldn't "rock
the boat" there, especially since Linux doesn't offer the same
capabilities.  No UNIX systems have without going with a very
proprietary solution, right?

As such ...

We shouldn't test these concepts at all.

Or if we test, we should _only_ test where they supplement or augment
systems or Linux capabilities around existing proprietary or
orphan/hostageware network/server solutions.

Especially since ...

Alan McKinnon  wrote:
> Actually, these industry people are being quite pragmatic. Like a
> site has implemented AD and would like to take advantage of Linux
> machines in cases where Linux has a proven track record. But they
> don't want to end up with two authentication systems and as AD is
> already there, they are asking "Please get this *nix box to use
> what I already have"

Again, most corporations (if not all) just deploy Novell eDirectory
or Microsoft ActiveDirectory.  And any such "open systems" solutions
are too difficult to maintain, too separate to do anything useful,
and there's nothing wrong with limiting open systems solutions using
enterprise Linux to areas where Novell and Microsoft is not so good.

And they shouldn't be tested -- if at all -- until ...

Alan McKinnon wrote:
> Tridge tells us that is still to come in SAMBA 4. Where I live,
> there is a real demand for that level of functionality
> [snip details of how ADS works]

Samba 4 comes out -- again, if at all!  That's when Samba will offer
an ADS-like, complete infrastructure system that will be compatible
with Windows 2000-era ADS, with possibly some 2003 compatibility.

But even then, Samba won't offer compatibility with Vista in its
native modes, and it won't work with Longhorn Server ADS and related
Microsoft 2007+ services.  So that really means we shouldn't bother
at that time either.

Thank you for educating me guys!
I am an idiot and I just see things wrong.

Etienne -- I was wrong about Samba being important.
Alan -- I was wrong about Samba not being important.

I was just totally wrong to think that LPI level 3 could focus on a
pure, open standards, perpetual enterprise network infrastructure,
including interoperability with older orphanware as well as newer
proprietary and even some hostageware systems.  Because I didn't
realize no major enterprise I've ever worked at has used anything but
Novell eDirectory and its stack of services and Microsoft Windows ADS
and its stack of capabilities.

Thanx for setting me straight!  I am an idiot, so I will "stop
spouting presumptions" on the LPI lists.

Bryan J. Smith   Professional, Technical Annoyance
b.j.smith at ieee.org    http://thebs413.blogspot.com
     Fission Power:  An Inconvenient Solution

More information about the lpi-discuss mailing list