slow.log by date
well, our slow.log is pretty huge and say I just want the logs in the last week. (tail -n XXX is not good enough) I'm wondering there is a script (Perl or bash or others) can do this? I'm not willing to reinvent the wheel. :) Thanks. -- Fayland Lam // http://www.fayland.org/ Foorum based on Catalyst // http://www.foorumbbs.com/ -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: slow.log by date
[EMAIL PROTECTED] wrote: Syslog-ng combined with logrotate well, logrotate is for the coming log. what can we do towards existing ones. :) Thanks. --Original Message-- From: Fayland Lam To: mysql@lists.mysql.com Subject: slow.log by date Sent: Nov 30, 2008 18:22 well, our slow.log is pretty huge and say I just want the logs in the last week. (tail -n XXX is not good enough) I'm wondering there is a script (Perl or bash or others) can do this? I'm not willing to reinvent the wheel. :) Thanks. -- Fayland Lam // http://www.fayland.org/ Foorum based on Catalyst // http://www.foorumbbs.com/ -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
which solution is better for $count and @cols
well, we have a where $where, and I want some @cols depends on $start, $rows. besides, I want $count too. so we have two solution here. A, two SQLs. 1, SELECT COUNT(*) FROM table WHERE $where 2, SELECT col FROM table WHERE $where LIMIT $start, $rows. B one SQLs with some operation SELECT col FROM table WHERE $where while $count is scalar @cols and real cols is splice(@cols, $start, $rows) which solution is better? or it depends on the $count, big count A is better and small is B? Thanks. -- Fayland Lam // http://www.fayland.org/ Foorum based on Catalyst // http://www.foorumbbs.com/ -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
Re: which solution is better for $count and @cols
Thanks very much.
D. Dante Lorenso wrote:
Fayland Lam wrote:
well, we have a where $where, and I want some @cols depends on $start,
$rows. besides, I want $count too.
so we have two solution here.
A, two SQLs.
1, SELECT COUNT(*) FROM table WHERE $where
2, SELECT col FROM table WHERE $where LIMIT $start, $rows.
B one SQLs with some operation
SELECT col FROM table WHERE $where
while $count is scalar @cols and real cols is splice(@cols, $start,
$rows)
which solution is better? or it depends on the $count, big count A is
better and small is B?
Use A always. You might get away with using SQL_CALC_FOUND_ROWS, but
I've always found that I need to know the total row count before I run
the query because if you are asking for a $start which is beyond the
$count, I want to modify $start before running the second query.
It ends up being like this:
A, two SQLs
1, SELECT COUNT(*) FROM table WHERE $where
1.5, if ($count $start) { $start = 1; }
2, SELECT col FROM table WHERE $where LIMIT $start, $rows.
Option B is horrible for large result sets. Only drawback to A is the
tediousness of having 2 queries, but you get over that once you
develop a pattern for writing them that way.
-- Dante
--
D. Dante Lorenso
[EMAIL PROTECTED]
--
Fayland Lam // http://www.fayland.org/
Foorum based on Catalyst // http://www.foorumbbs.com/
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
recommended SQL for 'category' and 'help' table
hi list. I want to build a category system and a help system. I think that's a very common task, and I'm curious about what's your SQLs to build these. I'm really appreciated your sharing the table structure with me, Thanks. Regards, -- Fayland Lam // http://www.fayland.org/ Foorum based on Catalyst // http://www.foorumbbs.com/ -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe:http://lists.mysql.com/[EMAIL PROTECTED]
