From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | "Jim C(dot) Nasby" <jnasby(at)pervasive(dot)com> |
Cc: | pgsql-hackers(at)postgreSQL(dot)org |
Subject: | Re: Max size of a btree index entry |
Date: | 2006-07-19 23:39:14 |
Message-ID: | 17034.1153352354@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
"Jim C. Nasby" <jnasby(at)pervasive(dot)com> writes:
> Could you not just scan right and see what the first key was? Thought
> granted, that means there's a chance of a wasted page scan, but I think
> that'd be somewhat of a corner case, so it might not be bad.
No, because (a) that confuses the first key that happens to be on a page
with its keyspace boundary --- what happens when you need to delete that
data key? and (b) because of locking considerations, you don't want to
move right and then have to back up. You'd have to hold lock on the
first page while reading in the second, which makes for a nontrivial
performance hit.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Tom Lane | 2006-07-19 23:45:24 | Re: Freezing tuples on pages dirtied by vacuum |
Previous Message | Jim C. Nasby | 2006-07-19 23:37:34 | Re: How does the planner deal with multiple possible indexes? |