Vertica?

[GrandMauler]

Ok, I have a senior manager making a push to a Vertica Database.

I have never heard of this database. I did a Wiki lookup and, while it does look prominsing in its querying speed, I'm not sure how compatible it would be with things like OLE Db connections and .NET assemblies.

If anybody has any info on this, please share it with me. I'm attempting to remain objective and will continue to do research.

But, off hand, I'm against it because I think it would basically mean completely redesigning our entire back-end infrastructure. I don't have to elaborate what a total nightmare that would be.


Thoughts?

 

[SQLSister]

My thought on this is that you should prepare a document analyzing the cost/benefits of moving to this database. Make sure to include an estimate of the number of hours it takes to design the database and to convert the exisitng data and front end code to use the new database. Make sure to qualtify this cost using the avergage developer salary times at least 2 (to account for benefits and other corporate support such as HR/admin).

Analyze the actual speed improvement. How slow is your current system, does it need resign? Could performance tuning fix the problem or faster equipment? Both would probably be significantly less expensive than completely resigning and migrating your database. Show these as options to fix the problems you currently have as well as the do nothing and use this new db options.

Pick a number of significant factors that you would evelaute any solution to your problems with. I suggest
cost, avaliablity of trained employees to make the change, likelyhood of introducing new, significant bugs to the system, time to make the change, likelihood to solve currently identified problems. I'm sure you get the idea and can think of some more. Then get the deciding managers to rate these decision factors by importance (1-5 with 5 the most important). (this is the key step!). Now rate each of your options (from 1-5 (5 is again the highest rating)) as to how well they would meet the rating factor (you do this not the manager). Multiply by the rating factor and add up the results. This will show the managers what is the best option becasue it will have the most points. Write a short document to support this explaining why you gave the ratings you did. keep it no more than two pages (managers won't read more anyway).

I have never failed to have management accept the option I recommended when I did this. They set the rating factor importance, so they can't argue with it. Very few will even try to change the ratings for each option, but even when they do, usually it isn't enough to change the outcome unless it was very close.

"NOTHING is more important in a database than integrity." ESquared