Scale-up vs scale-out - what are they?

Basically, “scaling up” means to add power to your server, or get a bigger server -build it up. “Scaling out” means to add new servers to your configuration - build it out.

Oracle touts both approaches, using RAC on big servers or on “blades”, or a “grid”.

MySQL generally touts scaling out, and most of the big sites using MySQL use “scale-out”. In my experience, and according to Jay Pipe’s excellent blog entry on the subject, that’s partially because MySQL doesn’t scale up as well as it scales out. (Another reason he mentions - that also fits my experience - is that people who pay for Oracle licensing match that by also buying expensive machines to put it on.)

In Oracle, whether you scale up or out, you’ll probably use RAC, which means setting up private interconnects and basically using “shared everything” with your servers. You’re still sort …

I just re-installed my MacOS 10.4 (Tiger) from scratch. However, it's more than just putting in the DVD and doing an upgrade.. This post is more a reminder for myself when I'll have to do it again!

For MySQL you'll need newer versions installed on MacOS 10.4 of the popular GNU tools: automake, autoconf, bison and libtool. Otherwise you'll notice it doesn't really compile. You'll find URLs here below and download the latest versions.
Additionally I needed to install pkgconfig because /usr/local/bin/autoreconf just failed for the ndb-connectors! Don't ask me why, I just found while googling for it..

I did install everything under /usr/local and changed my $PATH in ~/.bash_profile to PATH="/usr/local/bin:$PATH"

Quick "one liner" for shell to check on the version of the build tools:
shell> for p in automake autoconf bison pkg-config; do echo -n "$p: " ; $p --version | head -n1; done; glibtool …

I just re-installed my MacOS 10.4 (Tiger) from scratch. However, it's more than just putting in the DVD and doing an upgrade.. This post is more a reminder for myself when I'll have to do it again!

For MySQL you'll need newer versions installed on MacOS 10.4 of the popular GNU tools: automake, autoconf, bison and libtool. Otherwise you'll notice it doesn't really compile. You'll find URLs here below and download the latest versions.
Additionally I needed to install pkgconfig because /usr/local/bin/autoreconf just failed for the ndb-connectors! Don't ask me why, I just found while googling for it..

I did install everything under /usr/local and changed my $PATH in ~/.bash_profile to PATH="/usr/local/bin:$PATH"

Quick "one liner" for shell to check on the version of the build tools:
shell> for p in automake autoconf bison pkg-config; do echo -n "$p: " ; $p --version | head -n1; done; glibtool …

According to recent survey, over 20% of MySQL users plan to migrate a Microsoft Access applications to MySQL over the next 12 months. However there are few documents available that describe best practices for performing such a migration.
This document summarizes discussion from the ?MS Access Migration? session at the 2007 MySQL User Group meeting in California. That session brought together a number of MySQL users with a goal of identifying key success factors for moving MS Access applications to MySQL.

And the surprises are not over. More is coming. Stay …

Zmanda’s MySQL Backup and Recovery Solution is the only comprehensive backup solution that addresses the unique needs of MySQL databases — such as multiple database storage engines. Zmanda Recovery Manager (ZRM) for MySQL now features graphical tools to simplify the process of protecting MySQL data.

This free presentation from experts in open-source backup and recovery of MySQL will demonstrate Zmanda Management Console for ZRM , which is a web based service integrated with Zmanda Network. It provides management and visualization tools for defining all backup activities of a MySQL DBA:

•  what to backup, e.g. all databases or just the selected tables

• where to store the backup images

• when to perform backups, e.g. schedule daily, weekly or monthly full and incremental backups

…

There is no "first" or "last" aggregate function in SQL. Sometimes you can use MIN() or MAX(), but often that won't work either. There are a couple of ways to solve this vexing non-relational problem.

First, let's be clear: I am posing a very non-relational problem. This is not about the minimum, maximum, top, most, least or any other relationally valid extreme in the group. It's the first or last, in whatever order the rows happen to come. And we all know rows aren't ordered -- in theory. But in practice they are, and sometimes you need the first or last row in a group. This article shows how.

Last week I noticed a blog port by Morgan Tocker on "Big transactions
suck.."

http://mtocker.livejournal.com/30197.html

His post made me think a bit about Innodb's behavior both in cases
where user's use small numbers of transactions, and the case where
user's write applications that connect and then disconnect from the
database.

The connect/disconnect data came out as I expected it. Reconnecting
to the database can cost you in performance. In each of the cases
below the test was run with an increasing number of users.

The commit vs autocommit was bit more interesting. From the graph I
can't find any discernible pattern showing which might be better
(though this highlights that I need to add standard deviation to
mysqlslap so that I can tell if the …

Got back from the MySQL DBA boot camp in San Francisco. Very nice one-day class, covered a lot. Kudos to Ronald Bradford; he did a great job of it. Good to get a free copy of the MySQL 5.0 certification guide and free passes to the certification exams. Now I have no excuse not to get certified… A lot of it I knew already (from having done this for some months now), but I especially enjoyed the section on tuning, and got some good tips throughout.

One question came up that is to the point of this site: where should an Oracle DBA start? Check out Ronald’s excellent and thorough answer at his site, in his post, “Learning MySQL as an Oracle DBA“. Certainly could not have said it better myself.

I've been thinking about this today. My attempts thus far to write random queries which stress various parts of the server have worked, and served a purpose - but have been rather huge and clumsy.

For example a 5 union select, each comprising of 7 joins did have it's value.. The problem with huge random queries is simply that the results cannot be verified as easily as they were constructed (unless you run the same on various DBMS). These huge tests are mostly only good for checking if the server crashes.

Now, I wish to try write testcases that have these properties:

• small
• results are self-verified
• each test is random, but reproducible given a seed value
  

Tonight I'll throw together a code-generator to do simple insert/select and post any interesting results here later.