Re^2: Is bind useful for one-time queries and commands?

I'm confused .. your initial statement

Any query to a DB2 server must first go through a bind step. ...

contrasts with your closing statement

... At least that's how binding in DB2 works (to my recollection) with compiled languages. I don't know how DBI and interpreted Perl handle the bind step. It may very well be that what you call a bind step through the DBI interface may have nothing to do with the DB2 bind step.

Do you have experience with DB2 and Perl, or is this more of an educated guess?

Alex / talexb / Toronto

"Groklaw is the open-source mentality applied to legal research" ~ Linus Torvalds

Comment on Re^2: Is bind useful for one-time queries and commands?

Replies are listed 'Best First'.
Re^3: Is bind useful for one-time queries and commands? by ruzam (Curate) on May 20, 2008 at 19:00 UTC
The DB2 server will execute the bind step internally if it has not already been done by the application. In other words, DB2 will intercept it's own query requests and bind them (temporarily) before actually getting the data. I don't know if or how long DB2 might cache ad hoc query requests, it's probably configurable. A cached query request would effectively eliminate the need for a bind step at all from the application side. I have experience with DB2, but not DB2 and Perl. It's an educated guess.	[reply]
Re^4: Is bind useful for one-time queries and commands? by talexb (Chancellor) on May 20, 2008 at 19:10 UTC
Update: Sorry -- I did the benchmarks wrong. Please disregard. I have updated the thread in question, in case you are interested in a more up to date benchmark. OK -- I've run benchmarks and found that `bind` doesn't slow things down at all -- my test actually ran faster with `bind`, which is what I should have expected. Now I'm trying to understand whether the type checking that `bind_param` does is really useful. Alex / talexb / Toronto "Groklaw is the open-source mentality applied to legal research" ~ Linus Torvalds	[reply] [d/l] [select]