Discussion Locked

2-tier VS n-tier databases

0 helpful
5 replies
Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.

Here's a slightly more challenging question for you:

Can you explain to me the impact on the Total Cost of Ownership between 2-Tier and n-Tier (more than 2) databases in a distributed database architecture?

Please reference thin-client models versus thick-client models and where business rules would reside in these architectures.

If you can help me with this one, I would greatly appreciate it.....

arongirl@home.com

5 replies

by john.m1

Sorry, i don't even understand the question but,
considering your opening line "more challenging question" and closing line "greatly appreciate", may i with respect suggest that you upgrade the points that you are offering. This should attract more knowledgable answers.

by carollong

This would require a full report from a consultant! No space for that here.

Client server computing cost of ownership really depends on how it is implemented and if that really fits the business need.

You can assume that for every tier you add, you will split out another part of the system onto another server and this will require the associated middleware. It also may increase network traffic which will require more capacity. This adds costs.

If the application is well suited to being broken ups into bits, the development costs and the cost of renewing various bits of that development will drop.

For thin clients, business rules tend to be elsewhere. Very thin clients don't even check the data you put in - it all goes to a validation server (which may or may not contain business rules - they might be in the application server tier.) Thin clients often don't need such a high spec so tend to be cheaper.

Thick clients are more likely to be powerful machines (more work to do, more cost) and will do much of the basic validation server work and may have some very basic business rules (complex ones on the application server or as constraints in the database). This means you only send usable data once over the network - lower cost there.

This discussion is closed to new replies.