Paul DuBois [mailto:[EMAIL PROTECTED] writes:
At 15:18 -0700 8/20/03, Michael S. Fischer wrote:
This is trivial to benchmark yourself. Try:
BENCHMARK(10, SELECT yada, yada FROM test WHERE id IN(1,2,3))
And compare to
BENCHMARK(10, SELECT yada,yda FROM test WHERE (id =1 or
id
In my opinion, the discussion is moot.
Unless you're running MySQL in a read-only environment, you should never
restart the daemon automatically, because an improper shutdown is likely
to yield table corruption, and if the tables are corrupted, attempting
to write additional data can cause even
It's quite possible you're using the wrong tool for the job. Since this
is a write-intensive environment, you may get better performance by
using another database such as PostgreSQL or Oracle. Alternatively,
consider the option of re-architecting the application to distribute the
writes across
The database does write the data to disk when an UPDATE or INSERT is
sent, but not synchronously. (IOW, it doesn't call fsync() after each
write.) So there is no guarantee that the data will be in the tables
when a power failure occurs. This is a tradeoff MySQL makes for speed
because it's not
Keep in mind that Linux will allocate nearly all of its free RAM to the
buffer cache as the kernel opens and reads files to increase filesystem
performance (cat /proc/meminfo and look at the buffers row) So, lack
of free RAM may not mean what you think it means.
As for the too many
that, or
alternatively, you can establish two database instances and swap between
them, performing the defragmentation on the inactive database.
--Michael
-Original Message-
From: Jack Coxen [mailto:[EMAIL PROTECTED]
Sent: Wednesday, August 20, 2003 11:43 AM
To: 'Michael S. Fischer'; Jack
This is trivial to benchmark yourself. Try:
BENCHMARK(10, SELECT yada, yada FROM test WHERE id IN(1,2,3))
And compare to
BENCHMARK(10, SELECT yada,yda FROM test WHERE (id =1 or id = 2 or id
=3))
See http://www.mysql.com/doc/en/Miscellaneous_functions.html for
documentation on the
]
Sent: Wednesday, August 20, 2003 3:14 PM
To: Michael S. Fischer
Cc: [EMAIL PROTECTED]
Subject: RE: Need help optimizing query, awfully slow on only
2 records
Result of EXPLAIN is:
table|type|possible_keys|key|key_len|ref|rows|Extra
inquiries|ALL|contact_id| | | |8253
Sure, just shut down the database cleanly (mysqladmin shutdown), move
the data directory intact to the new filesystem, and start up again.
--Michael
-Original Message-
From: Jon Miller [mailto:[EMAIL PROTECTED]
Sent: Wednesday, August 20, 2003 8:41 AM
To: [EMAIL PROTECTED]
Subject:
You're going to need to architect more than just a master-slave
relationship to do what you want to do. All replication does is copy
commands from one MySQL server to the other; it does not create a
failover environment by itself.
Creating a failover environment is beyond the scope of MySQL;
What does EXPLAIN SELECT query show? Have you read the chapter in the
manual on optimizing queries? Do you have all the proper indices set
up?
--Michael
-Original Message-
From: Apollo (Carmel Entertainment) [mailto:[EMAIL PROTECTED]
Sent: Wednesday, August 20, 2003 2:52 PM
To:
In a word, no. The way MySQL organizes its datafiles is trivial by
comparison: one directory per database, two files per table (table.MYI
and table.MYD), one is the datafile, the other is the index file. MySQL
also does not preallocate space for its tables like Oracle does.
--Michael
Michael,
The reason you think there's not much to this database is that compared
to Oracle, there really is not much to this database. :-)
From a 50,000-foot point of view, it's really just a nice SQL interface
to ISAM files (it started out that way, anyway). It is designed for
Are you connecting through a TCP or a UNIX domain socket?
--Michael
-Original Message-
From: news [mailto:[EMAIL PROTECTED] On Behalf Of Jon Drukman
Sent: Tuesday, August 19, 2003 2:36 PM
To: [EMAIL PROTECTED]
Subject: PHP mysql_connect randomly failing
I've got a library of
The MySQL client won't be able to do that on its own; you'll need to
write a script in your favourite scripting language to do that.
--Michael
-Original Message-
From: Scott Haneda [mailto:[EMAIL PROTECTED]
Sent: Tuesday, August 19, 2003 9:00 PM
To: MySql
Subject: marking all
I'm using MySQL 4.0.13 for a web application, and one of my queries is
sufficiently complex such that it might take several seconds to return
the results to the user. I'd like to return a please wait page if the
query is not already in the cache. Is there a way (or a proposed way)
of doing a
Jeremy Zawodny writes:
Perhaps you could tell us more about the queries that take
too long. Maybe there are some things you can do to reduce that time?
You don't want to go there.
But because you do... :-)
As far as I can tell, the queries are as optimized as they're going to
get. I'm
My InnoDB file is set to 60MB, and is not set to autoexpand.
When running stress tests against my server today, I got a message that the table was
full and the test stopped. Sure enough, the InnoDB file was about 62MB.
My question:
I assume that the InnoDB file is the transaction log. Is there
: Michael S [EMAIL PROTECTED]
Reply-To: [EMAIL PROTECTED]
Date: Thu, 3 Apr 2003 23:45:28 -0600
My InnoDB file is set to 60MB, and is not set to autoexpand.
When running stress tests against my server today, I got a message that the table was
full and the test stopped. Sure enough, the InnoDB
19 matches
Mail list logo