The Best Oracle Parallel Execution Paper

I'd already made a note to read and blog about Greg Rahn's Parallel Execution over RAC blog post which is excellent, as usual, but Jonathan Lewis beat me to it What struck me when I eventually took the time to read it, though, was the link in his post to an Oracle White Paper (PDF) on Parallel Execution which I hadn't seen, despite it being published last year? When Oracle do get round to publishing White Papers expanding on the documentation, I typically find them superior to a lot of community efforts (much as the documentation is still usually my first port of call). This is no exception.

It's terrific stuff because it combines more technical rigour than my own papers based on my own experience with several statements that reflect my real world experience very well. Or, to put it another way, it's technically sound and detailed but never loses sight of the day to day problems that people face. I could have picked out tons of quotes from it, but here are a couple ...

"Do not use a single 1 TB disk for your 800 GB database, because you will not get good performance running operations in parallel against the database; this might work well for your single-user home video archive, but not for a database leveraging parallel query with multiple users."

"The internal code rewrite introduced with Oracle Database 10g lifted a number of parallel execution restrictions that existed in Oracle Database 9i. As a result you might see that some operations that were running in serial are now executed in parallel when you use parallel settings at the table level. This may be great for the execution time of these operations that did not run in parallel before, but it also means that the system will end up using a lot more parallel resources than it used to."

Whilst I don't need an Oracle White Paper to confirm what I've seen in the field, it's nice when it does. Oh, and I'll say nothing about what it has to say about parallel_adaptive_multi_user other than to say I think it's clearer than anything I've seen coming out of Oracle in saying it might not always be a good thing

So, yes, it's better than anything I've written on the subject but the unfortunate side-effect is that my plans to update the PX papers very soon can be delayed Actually, I wish they could, but as I've just had two abstracts accepted for UKOUG 2009, one of which is titled 'More Parallel Fun' it might be clear that

a) I was thinking about and have been doing more PX work anyway; and
b) I can't get out of it now

Seriously, that paper might be quite long, but it's well worth your time reading it if you're remotely interested in the subject. Oh, but if you're going to print it out to read it, make sure it's a colour printer! The use of colour is effective if you know about it in advance

P.S. One more thing. It's yet another good paper to come out of Oracle that could have done with one more proof reading (spelling here and there), but I am enormously picky at times - usually with other people's stuff!