why locking a path in B-tree is not necessary
From: Mikito Harakiri <nospam_at_newsranger.com>
Date: Sun, 09 Sep 2001 22:01:14 GMT
Message-ID: <KERm7.1440$%u4.1368_at_www.newsranger.com>
I must apologyse in advance for posting a question, the answer to which I may easily find, for example, in Gray/Reuter's book.
Date: Sun, 09 Sep 2001 22:01:14 GMT
Message-ID: <KERm7.1440$%u4.1368_at_www.newsranger.com>
I must apologyse in advance for posting a question, the answer to which I may easily find, for example, in Gray/Reuter's book.
There is no question that [physical] denormalization like indexing, or creating materialized views decreases the degree of concurrency. So, even though, the individual query might speed up, the total throughput would suffice. The question is how much?
Received on Mon Sep 10 2001 - 00:01:14 CEST