Wednesday, April 16, 2008

What I learned yesterday...

My boss gave a few of us in IT a project management crash course to prepare for a year+ long integration of a large acquisition. I will be involved with the networking team but this was a more general course so we'd all understand the big picture and process. I learned what a CRP is first of all. Conference Room Pilot: basically the stage where you are doing everything in a test environment to ensure it is perfect before moving forward to implementation. It is more of a process because you end up having several CRP meetings with a final CRP which must be signed off on by the top level steering committee.

Also I learned about the Triple Constraint Law of project management. Basically it says there are three general aspects to any project plan: Time, Cost and Quality/Scope. These goals can be stated in plain English: Fast, Cheap and Good. The law goes on to state that only two of the three of these goals can be met in any project.

In other words if you decide you want the project results to be good and you want the project done fast... it will NOT be cheap. If you want the project results to be good and you want the project to be cheap, then it will NOT be fast. Lastly, if you try to do the project fast and cheap... you guessed it; it will not be GOOD!

It was a surprisingly interesting subject to me. Perhaps one of these days I will go to night school and get a MBA. After all, I don't imagine I will want to have to learn the latest Networking technology when I am 60 years old... with a MBA I can just move on up to management and boss other people and let them learn the ins and outs of IP version 8.


The Irascible Neufonzola said...

Hmmm, good point on the MBA. I've long thought about that sort of thing, but never very seriously...but the constantly learning new technology thing might get a little old after 20, 30 years.

But re the triangle, I learned a similar concept, except with a square. In each corner is one of the following:

* Features
* Quality
* Cost
* Time

It is focused on software development, but can be applied to other things. The only real difference with yours is that it adds "features", which I guess you could clarify as quantity of features. You can, at best, only optimize two of these at a time. Meaning, you're doing good if you're able to get a full featured product out quickly, but be prepared for bugs and a high bill. Likewise, you might be able to create a project that is bug free for cheap, but you won't be able to do it quickly, and it will be very simple by necessity. Any two of those can be optimized for a project at the expense of the other two, and any single one can be maximized at the expense of the other three.

That sort of stuff does interest me. Maybe because I'm used to a manager who wants a full-featured software package delivered "yesterday" without bugs, and for dirt cheap. As Sir Mick intoned, "ya can't always git, whatcha want...but if ya try sum time, ya jes might find, ya jesjesjes might find, ya git whatcha neeeeed!"

Percussivity said...

You must be joking... Mick Jagger was knighted????????

Those silly brits.

The Irascible Neufonzola said...


The Angry Coder said...

Welcome to the real world of corporate IT!

The Irascible Neufonzola said...


You're the Exchange 2007 guru, now, right?

Can Exchange 2007 be loaded on a x64, dedicated Server 2003 box, in a domain in which the DC is a SBS 2003 box? To replace the "embedded" version of Exchange 2003 on the SBS machine?

I ask because that was my plan, but the vendor/IT consultant I am dealing with is now telling me that "can't" be done. I produced a number of articles to the contrary and he amended his statements to that "shouldn't" be done.

I don't have a budget for replacing that SBS box completely with a "real" server at this point. Too many technologies still in use on it, and too much that would have to be reimplemented. But I really, really want to get Exchange off of that box and give it a dedicated server box. And due to the very poor advice of the aforementioned IT consultant, we have a $6000 x64 server sitting around doing all of nothing, along with a copy of Exchange 2007 from an Action Pack subscription. So...I'd love to make use of this stuff, but since my trust level with the consultant isn't all that high, I'd love to get a second opinion, if you've got one.

Percussivity said...

Actually no I am not as of yet the Exchange 2007 guru... maybe this time next year.

I do know that 2007 is natively 64 bit and MUST be installed on a 64 bit box with a 64 bit server OS. Are you saying you want to remove Exch2003 from the SBS and then do a clean install of 2007 on the other x64 server?

I wouldn't see a problem with that unless removing 2003 will do something nasty to the SBS. It could be like when a person gets an alien embryo inside of them and it is impossible to remove without killing the host.

The Irascible Neufonzola said...

Yep that's what I'm thinking. You're right, I have a dedicated 2003 server in x64 (it was destined to be a SQL server, but we only had x32 SQL available) that is doing nothing productive now. We tinkered with the idea of making it a high powered Terminal Server, but man, client apps on a x64 box are a pain in the neck! So Exchange 2007 sounds ideal for it. And I've got a license and 10 CALs already, I'd just need to buy the remaining CALs. My only concern is how Exchange is integrated with AD...does the Exchange box have to be a domain controller? And if so, can a Server 2003 box be a slave DC to a master DC on an SBS 2003 box? Hmmm.

Percussivity said...

It doesn't have to be a domain controller... but if it is then I believe it must be a global catalog.