CS/TP Performance: 3-tier vs 2-Tier?

CS/TP Performance: 3-tier vs 2-Tier?

Post by denn.. » Sun, 29 Jan 1995 23:09:38



Where can I find solid PROOF that 3-tier CS/TP apps have superior performance?

All indications are that Client/Server OLTP applications built using a "3-tier"
architecture scale significantly better than "2-tier" applications.  When the
number of users is 100's or 1000's 2-tier apps will "hit the wall" but 3-tier
apps will keep on going.  We know this from the standpoint of systems arch.
and *limited* internal benchmarks.  But we'd like to be able to pin it down.

Can anyone help me find proof, in the form of:

    1. Benchmark results
    2. Documented real-world experience / war stories
    3. Anecdotal evidence (as a last resort)

This is very important to my company.  Requests for anonymity will be
respected.  All help greatly appreciated!

--

 
 
 

CS/TP Performance: 3-tier vs 2-Tier?

Post by Jeff Sutherla » Sun, 29 Jan 1995 23:32:15


Quote:>Where can I find solid PROOF that 3-tier CS/TP apps have superior
>performance?

>All indications are that Client/Server OLTP applications built using a
>"3-tier"
>architecture scale significantly better than "2-tier" applications.  When
>the
>number of users is 100's or 1000's 2-tier apps will "hit the wall" but
>3-tier
>apps will keep on going.  We know this from the standpoint of systems arch.
>and *limited* internal benchmarks.  But we'd like to be able to pin it
>down.

>Can anyone help me find proof, in the form of:

>    1. Benchmark results
>    2. Documented real-world experience / war stories
>    3. Anecdotal evidence (as a last resort)

>This is very important to my company.  Requests for anonymity will be
>respected.  All help greatly appreciated!

Talk with Mark Winters Associates in Toronto.  They can build a $600K
Powerbuilder application for $100K in Smalltalk.  It runs not only faster,
but better, with more functionality.  And the users will like it better.

Jeff Sutherland
"C++ runs slightly faster than Smalltalk but a lot later!"


 
 
 

1. CS/TP Performance: 3-tier vs 2-Tier?

Where can I find solid PROOF that 3-tier CS/TP apps have superior performance?

All indications are that Client/Server OLTP applications built using a "3-tier"
architecture scale significantly better than "2-tier" applications.  When the
number of users is 100's or 1000's 2-tier apps will "hit the wall" but 3-tier
apps will keep on going.  We know this from the standpoint of systems arch.
and *limited* internal benchmarks.  But we'd like to be able to pin it down.

Can anyone help me find proof, in the form of:

    1. Benchmark results
    2. Documented real-world experience / war stories
    3. Anecdotal evidence (as a last resort)

This is very important to my company.  Requests for anonymity will be
respected.  All help greatly appreciated!

--

2. Free Books - honest!

3. 2-tier vs 3-tier architecture

4. ATL object - How do I override QueryInterface

5. CS/TP Performance: 3-tier

6. VS.NET with 400 mhz/4GB/64MB

7. from 2-tier to 3-tier

8. MPC107 memory boundary registers

9. 2-Tier, 3-Tier & CORBA Client/Server application

10. 2-tier & 3-tier

11. 3-Tier CS/TP (5:7) Flexibility

12. 3-Tier CS/TP: Data Dependent Routing