I have a bunch of tuning questions, and I would appreciate any answers to
any of them, in any order... :) Thanks in advance.
1. Clustering: Are the current docs corecct in saying that once
you cluster on an index, subsequent inserts will _not_ respect that
cluster? I only ask because a recent posting seemed to imply otherwise.
2. Speed difference between trigger functions in C vs PGPLSQL: Given that
I will need to use identical SQL inside the C function as in the PLPGSQL,
is there a big speed difference?
3. Query Execution Times: How do I find the number of seconds that a query
takes to execute? In MySQL, that info is printed after the results.
4. Inserting w/ Indexes vs. Selecting w/o Indexes: Given an equal number
of inserts and selects on a table, (a whole lot of both--about 10/second),
is it better to take the penalty on inserts due to indexes or on selects
do to lack thereof?
5. Does PG have partial indices yet? I only saw a little about this in the
docs.
6. Does PG cache inserts when using Perl, when using `$statement_handle =
$dbh->prepare( "insert into foo (bar) values (?)");
$statement_handle->execute();'?
7. How can one do If-Else chains in PL/PGSQL? If I am being braindead
about this one, please ignore it.
8. If I get good answers to this, should I patch the FAQ and send it in
for approval? Is there any other way to help (besides writing C, which I
don't do very well...)?
--
Webb Sprague
Programmer
O1 Communications