Re: kaio

In article <5p9q9o$klf@cssun.mathcs.emory.edu>, Clem Akins <clema@informix.com> writes >Clem pontificates: >>>But I have personally done a benchmark on a large OLTP system on HP and >saw a >>>10% reduction in average transaction time due to changing only KAIO. >>> > >David Williams laments: >> Why? Since KAIO threads run in the same process as sqlexec threads >> (which in effect "generate" I/O requests). >[...] >> I'm confused unless KAIO is poorly implemented in some way.... > >I think you are just confused. "Reduction in transaction time" means a >performance *improvement.* All the things you listed are correct, and the >KAIO made things faster for me for all those reasons. > >Clem Sorry, I put my reply next to the wrong part of your message... >I heard from an Informix tech who personally did his own BM and found the >performance >breakpoint was at about 3,000 concurrent (simulated with Tuxedo(?)) users, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >after >which KAIO significantly hurt performance. Up until that point it helped, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Why would "KAIO significantly hurt performance" above 3000 users? >______________________________________________________ >Clem Akins (aka clema@informix.com) >Informix Software, Inc (Standard disclaimers apply) >International Technical Support >Last seen: "Get back to work, you!" in Auckland, NZ -- David Williams