user12288492 wrote:Without re-running your tests EXACTLY, I couldn't tell you which I/Os relate to cleanout effects, what relates to block chains, and how the counts are affected by pins. What I can tell you is that you're worrying about a detail that might be completely irrelevant in a production system. An earlier post complained about the logical I/Os being much more than expected - but it was one more on the first access to the row by the first session to visit that row after it was created.
It means that now it has to look for id = 1 in two blocks . So I should be getting cr=2.
Why an extra I/O now? ( I had committed after inserting record for id=6)
rp0428 wrote:Sorry for the confusion - I should have been less vague.
So for the above definition of Modulo that uses quotient/remainder the remainder is the block and the quotient is the row within the block. Your use of 'remainder' for the row within the block is what was confusing.
I am sorry for all that. In fact, I did not know how to get it reviewed before publishing it.
As per your advice , I have deleted my post. I have edited it to reflect all the changes you suggested. How can I send it to you for your feedback?
Edited by: user12288492 on Mar 20, 2013 9:51 PM
I sincerely apologise for what has happened. In fact it was all unintentional. I never wanted to claim the contents of my post as my original work.
I am just a nobody without even elementary knowledge of oracle. It was just that I didn't know how to go about it.Well, we are all here to learn. I can speak about myself that I know nothing except just about 0.00001% of what Oracle db is. But the important things are two. One, make sure you try to learn as much as possible and two, be open to get corrected whenever you are wrong.