Sunday, May 08, 2005

The LANL Enterprise System

Numerous people have sent the following in reference to the LANL enterprise project. It has been repeatedly suggested that the LANL EP is a candidate for investigation regarding the costs of its development and value provided, and in fact it is listed on the "Running List of Wasteful Activities" sidebar link.

One comment listed there:

"Allowing a project such as the enterprise project to spend $200 million to install a system that will not deliver what it has promised to deliver. Giving that same project carte blanche to pay astronomical fees for contractors who have already proven that their greatest expertise is their ability to justify why they didn't deliver."

Thats the first thing I thought of when I opened the funny pages this morning. Scott Adams must be lurking around the lab.
Even funnier when you show puns such as this to people that are actually heavily a part of the Enterprise Project they laugh and shake their heads right along with you.
Dilbert's comic this morning was right-on almost to the $.
An effective LANL Management Information System (MIS) is about two decades overdue, but the Enterprise System is likely to be another abortive, and costly effort, per GAO reports . The reason behind it, of course, is the profound incompetence of the Computing Division management, where the leadership has been passed around within a clique of Good Old Boys for about two decades.
During the 1993 reorganization the Division leadership was competed, in a very limited way, against outside candidates, and two of them finished ahead of the LANL contenders. Donna Crawford, of Sandia, was one of them. Word is that Sig turned her down. Donna went on to have a nice career. The other candidate, from Cal Tech, turned down an offer, where upon Sig turned to Dayem, the incumbent, and another 12 years of incompetence ensued. Having established how readily one could turn up candidates superior to LANL contenders, a nation wide search would have seemed more appropriate. Not a brilliant move on Sig's part...
There is no reason, except incurable optimism, to think these people, famously incompetent, will suddenly produce a wonderful MIS. The shortest path to a workable MIS scheme is to remove UC. The other bidders have effective MIS schemes, and can install them here. Two decades of futile LANL schemes should come to an end.
We are left with the question of what the current UC/LANL management would do with a workable MIS scheme. Ignorance is bliss... in the UC dream world.
The earlier comments seems to reflect a profound misunderstanding about who is running the Enterprise Project.
We knew EP was crap right from the start when Bob Newell was involved. He lasted what? 3 months? All we've ever been told that EP has accomplished is a new T&E entry system... real value added.
CCN does not run the Enterprise project. IM only 'runs' it in name only. The people running it are very much not connected with any of the Computing Divisions.

One thing that future directors, managers, and contract companies should know is that a lot of 'Fact about XYZ group' at the library is usually based on a story from 10-15 years before. Some slight or war with someone who retired 10 years ago is still used to paint the current division holders as the blackguards. There have been multiple times where I have talked to someone in T division about problems in CCN, and get the feeling that the story being told was last week and people made Saddam's agents look like Boy Scout poster children. Asking how long ago it occurred turns out that the problem was back when CCN was called C and none of the people involved have been at the lab since 1994.

So when you are told about some problem with MST, NMT, X, LANSCE, CCN, etc.. find out how long ago the story really occured. Find out if the people and practices have happened, and find out if the problems really still exist. And if they don't please educate the original group nicely that things have changed and they should give it another chance. And if they dont want to. please let them find jobs elsewhere.
Rather than re-invent the wheel with some sort of enterprise project, why not simply use the available tools: Something like an internal blog.

There are plenty of proven-commercial products that could support this system. If you're looking for assistance, might I encourage the vendors to weigh in. They're reading this.

The ones that come to the table with low cost solutions are the ones more likely to get asked again.

Remember, don't try to automate a communication problem; need to first get a firm undrestanding of what broke down. Changing one man will not change managements tight-grip on informaiton.

I go into some examples at the link of various types of ideas that could be discussed in an internal LANL blog: In a way that would share information, get people thinking about issues in advance, as opposed to the last-minute reactions.

These communication issues need to be ironed out before start trying to automate what is self-evidently and ineffective information flow within the lab.
The 5/08 12:05 poster sure sounds like Chris Michels. This poster has his information all wrong about everything except the fact that Dayem was indeed chosen CIC Division Director after two excellent external candidates, Donna Crawford and Paul Messina, turned the job down. The information given on the EP project is totally skewed. The key thing to remember is that the original EP project, started by John Brown and headed by a former CIC division director was effectively squashed by GPN who had the project leader replaced because Pete "knew how to implement an ERP project" since he had done one for the Navy. Consequently the project is way behind the orininal schedule and, from the sounds of it, way over budget.
Chris Mechels needs to get a life!
Lest we forget, remember who ordered the EP. Yep good old Nanos. Leaving before the depth of the mistakes and problems with it are understood.
Nanos, who?
To the 6:27pm commentator: To be accurate, John Brown ordered the EP project (called ERP back then) be implemented, not Nanos. When Nanos took over he sacked the original EP project leads and brought in his own team.
Let's see:

Joe Salgado started the EP based on a Gartner Group recommendation headed by Bob Newell. Charlie Slocumb ran it, then the bald guy Mike Something from Las Vegas who worked at LANL and IBM, then Nanos put Newell in charge for "the big bang." Newell was replaced by George Hansrote, a construction manager, by Rich "I'm starring at your chest" Marquez and Zirkle.
But Heidi Hahn is also involved: has she ever finished anything?
IM and Mousseau and the dynamic CIO duet of Neff & Lindsey have also been involved.
No wonder nothing has gotten done and EP is behind schedule.
Call the GAO and IG - talk about waste, fraud, and abuse !!!!
To 7:40am post:

It was Mike Payne.

IBM was replaced by ASRC and millions went to IBM and Oracle for consulting.

Talk about draining the swamp of $$$$
I remember Joe Salgado at one of his all hands: "I don't know what ERP stands for, but we should do this."

Another loser hired by LANL. He's living in Pojaque and is suing the Lab. I guess is proves you can sue LANL for its ignorance in hiring you.
I think CCN and other groups would have an easier time working with EP if some of the people dont come into meetings with lines like: "In 12 months you will be working for us." It might be true.. but it is not the way to influence people.

It would also help if the ERP system sold the reasons why it would be doing:
1) All lab mailing lists (internal and external).
2) All lab mail (pop, imap)
3) All lab authentication (well at least to Windows boxes.. others will have to switch as the presenter said.)
4) All lab desktop configuration (well for windows.. all others will have to switch..)
5) And only use Oracle approved security patches on their boxes (IE 5.0 unpatched because the later versions do not work or are not ok'd by Oracle.)

Not saying these arent things that should be centralized or not.. but better presentation would be key.
However, the market in the “real world” is willing to pay $100 per hour (and more) IF AND ONLY IF it is getting what it is paying for. No market (including LANL) should be willing to pay “$200 million” for George Hansrote’s and Linda Lambrecht’s failed management of this project and for their mix of Oracle consultants and Oracle software that have proven to be no better (and probably worse) than any of their predecessors.
And don't forget Heidi Hahn either as the third member of the EP Trinity.

Is Heidi's divorce final? Or is she just separated from Phil Goldstone?
and anything IM gets its grubby hands on is always a disaster since it's always looking for ways to justify its existence even tho is doesn't possess a crumb of management talent.
Ah, EP...the project that keeps on taking! What started as a 3 year 72 million dollar project with a fairly defined scope and delivery schedule has turned into a monster that devours money and souls at an alarming rate. Current estimates place its costs over $180 million dollars to date. And we have seen, uh, not much!
What I personally find hilarious is that the total allocation to expand LANL's Red or Classified Computing Network is about $20 M.
EP...$200 M
Classified Computing...$20 M
At least we got our priorities straight. Oh, and while we are at it, let's remove Kris Kemper and Richard Kendall, two of the best managers the lab has, and replace them with....a former EP project manager!
No wonder I spend so much damned time in the bar!
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?