New versions of MySQL and PHP motivated us again to build new versions of XAMPP. First of all we now could announce the release of an up-to-date BETA of the Linux version of XAMPP.

New in this BETA: MySQL (5.0.16), DIO support for PHP5, phpMyAdmin (2.6.4-pl4) and PHP (5.1.0). Warning: eAccelerator currently doesn't support PHP 5.1.

Follow me to the XAMPP BETA area

[b:3mphntqn]Update November 28th 2005:[/b:3mphntqn] The BETA now contains PHP 5.1.1 [b:3mphntqn]Update December 2nd 2005:[/b:3mphntqn] The BETA now contains also Apache 2.2.0

As reported earlier I have been taking the MySQL core and pro exams. This morning I found two letters from Denmark in my mail, wherein The Company deemed me fit and bestowed upon me the vast powers of the Certified MySQL Professional.

My certification IDs are 73899 and 74024 and the password is "public".

Getting into a support type role with some developers using MySQL. One particular developer wanted to use the "LOAD DATA INFILE..." command to load some data from a text file.

I think to myself, "No problem, just grant the FILE privilege".

He tells me "LOAD DATA INFILE" still doesn't work. After a little head-scratching, I pull out MySQL by Paul DuBois and realize it only works from the server and not from the client. Doh. (Although the "Access denied" error was pointing me to the MySQL maze known as privileges).

The behaviour of SYSDATE() has changed between MySQL 4.1 and 5.0. The manual indicates: Within a stored routine or trigger, SYSDATE() returns the time at which it executes, not the time at which the routine or triggering statement began to execute. This differs from the behavior for NOW(). Obviously, this was implemented internally by having SYSDATE() ignore SET TIMESTAMP, as can be easily tested.


Continue reading "sysdate() weirdness"

In a Perl application, using the widely known DBI module, a database connection is fired up by an instruction such as this one:

my $dbh = DBI->connect("DBI:mysql:test",
  "user_name", "my_secret_password", { RaiseError => 1 })
or die "something went wrong ($DBI::errstr)";

The first parameter is compulsory, and it's made of "dbi", which is the same for all connections, the database driver, which in our case is "mysql", and the database name, in our case "test". For a self respecting database, we should also specify the user name and password.
Nothing to complain about if this script is safely stored in your home directory, screened from prying eyes by the OS permissions.
But what happens when you need to distribute this script to somebody? You need to remember to change username and password to some dummy text, to avoid the burden of having to change our real credentials if by dumb chance we …

Poking around the Sun site tonight and stumbled into the announcement from last week about Sun's inclusion and support of PostgreSQL (or Sun Postgres) in Solaris 10. Yes, I know that throwing PostgreSQL and MySQL into the same sentence can get a person into trouble. I've had my fair share of confrontations . . .

There are a lot of questions raised by this. Right now I'm wondering about the support expertise:

Today Sun announced that it will be integrating the Postgres open source data base into the Solaris 10 OS and providing world-wide 24x7 support for customers who wish to develop and deploy open source database solutions into their enterprise environments.

I'm far removed from the PostgreSQL community, but the announcement from Postgresql.org …

What an evening! Swisscom experienced some major internet connectivity problems tonight, heavily affecting our MySQL servers. I was hacking away on my laptop for hours while my date was sitting next to me in front of the fireplace (until she left after some time). But Perl and MySQL once more saved my life (albeit not the romantic evening). I'll give you an overview of the events, maybe there's something to learn from this for others as well.

I got the phone call from the office at 20:00 CET while at the same time my mailbox was filling up with messages like this one:

Subject: jobRunner@linuxautomat2 - Error executing job 'METAR Archiver'

Error: './metarArchiver.pl -c archiver.conf -aq' returned exit code 22
Output from './metarArchiver.pl -c archiver.conf -aq' (STDERR):
Error: Couldn't open mysql db 'meteonews' on 'mysql1.intern'
       (Too many connections)

My workmate in the office first blamed the database to be the …

There has been a lot of press coverage of MySQL recently, since Oracle’s purchase of Innobase Oy (all your Innobase are belong to us hahaha) in early October, and the recent release of MySQL V5.

Here are links to some of the stories I’ve read

• MySQLs announcement. (link to PDF)
• Charlie Garry’s opinion on Oracle’s Innobase purchase.(link)
• Lisa Vaas’s article on Oracle’s purchse. Lisa suggests MySQL can get back together with Sleepycat and work on an enhanced bdb engine type. One item in the story stands out as flawed; Michael Stonebraker’s quote: “If I were the MySQL guys, I would be terrified that that engine was owned by somebody …

If you run the following program as a non-root user, you will see exactly how much memory a single process can allocate on your Linux system. The program is simple, it gets memory in 1MB chunks until it is shot down by the Operating System. In order to prevent overcommitment, the memory allocated is actually written to with memset().

On a Debian 2.4 smp kernel with 4 GB of RAM, it stops at 2933 MB.

(via Torsten Sievers)


Continue reading "How much memory can a process use?"