Brian, As your volunteer strap-on conscience for the lists, I thought I'd mention you're getting out of the technical discussion area and into your patented "if only the fools had listened to me" mood again. (Grin).
Ross Sent via BlackBerry by AT&T -----Original Message----- From: "Bryan J. Smith" <[EMAIL PROTECTED]> Date: Fri, 9 Nov 2007 10:31:04 To:"This is the lpi-examdev mailing list." <[email protected]> Cc:"Bryan J. Smith" <[EMAIL PROTECTED]> Subject: Re: [lpi-examdev] LPIC1 obj. update [EMAIL PROTECTED] wrote: > I would agree with Brian. This certification is for enterprise use. My favorite example of this difference is when approaching directory services. If you ask Linux users how they setup directory servers, at least 97% of them will tell you to setup Samba to connect to Windows ActiveDirectory Server (ADS). I know this thought dominates much of the discussion on this matter. The truth of the matter is that many enterprises have a separate, peer LDAP/Kerberos realm from ADS for UNIX/Linux. Linux in the datacenter is this reality. Yes, there are sometimes peer replication between them, or even shared/delegated Kerberos realms/principals, but they are separate. The point here is that we test for the open enterprise, not the Windows enterprise. I piss a lot of people off with that, but it's reality. There is a reason I have not associated myself with LUGs and other "community" lists (except development) for most of the last 3 years. I am regularly made fun of for my posts (which, ironically, is why I got most of my work when I was an independent -- from those same "no one agrees with me" views ;). And one too many times I've had to save people from other people screwing up their network. In fact, this drives most of my posts, because sometimes you can't avoid telling someone, "no, don't do that, it works for web, but not what you're doing, you'll screw up their network." And after a few more posts, I quickly realize the person suggesting such has never worked in such an environment. They eventually call exposing that a "personal attack" -- even though, in more than one case, I actually flown out (out of my own charity/expense) to the person's site to save them from their implementing someone else's solution. I see that on the LPI lists regularly. People tooting off what they know, when they actually don't for the context under discussion. File locking is a huge one I run into regularly, with web approaches hacked into SMB/NFS, Oracle/other SQL, etc... Stuff that corrupts data in milliseconds outside of read-only, quick open/close HTTP-type access. [ I was going to post some ore background on my views, but it's hard to do that without several people thinking I'm "bragging" so I removed it. ] So while I'm all for "foundations" in LPI, make sure they are the "differentiating" from many other training and approaches -- things that enterprise not merely want, but feel other training or certifications don't address. If you want to make LPI a home/web-focused cert, that's fine, I'm not going to be a good SME. > Home environment is a good training tool. However there is a hole > different world when running Linux in a production environment. How > often does some have to deal with 2TB database with 3000 users > doing data entry or queries at home. > Or would have to deal with an application that would cost more then > twice your car. > I had to deal with problems with I/O bottlenecks when I had to work > with RHE when it first came out. > Basic concepts can start with home power users, but the knowledge > with dealing with enterprise production environment is what > companies are looking for. > What knowledge would home user have when they are interviewed and > the example is that they "interviewer" has an 8TB Oracle 9 > database, with JDE, and PeopleSoft that needs to upgrade to Oracle > 10G. Storage switching from file to raw. > 1.What backup strategy could would work in a windows, unix, and > linux environment? > 2. What methods for archive or offsite would be used. > 3. Would the existing infrastructure be supported or would it also > require an upgrade, and why? > Well off my soap box. Thats my thougts. As long as they can be put into objectives which map out to real-world tasks of the lower Bloom's Taxonomy level, they are valid considerations for a LPI exam. We also have to be the "lowest common denominator" that works across LSB compliant distros. That's a tall order. -- Bryan J. Smith Professional, Technical Annoyance [EMAIL PROTECTED] http://thebs413.blogspot.com -------------------------------------------------- Fission Power: An Inconvenient Solution _______________________________________________ lpi-examdev mailing list [email protected] http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
_______________________________________________ lpi-examdev mailing list [email protected] http://list.lpi.org/cgi-bin/mailman/listinfo/lpi-examdev
