Pythian Group

Welcome to week 2 of Blogrotate. It was a short week due to Thanksgiving (Canada) and Columbus Day (US), but the world of IT is always buzzing. So as they say at the race track, pitter-patter, let’s get at ‘er.
Read the rest of this entry . . .

Welcome to the 160^th edition of Log Buffer, the weekly review of database blogs.

MySQL

Blame it on MyISAM, says Mark Callaghan of High Availability MySQL, on considering sql_mode and type coercion. “I think that MyISAM has its place,” writes Mark. “It does fast table scans, but InnoDB is much faster on just about everything else. I am just not thrilled with the impact it has had on MySQL.”

Not that those other engines are without flaw. Peter Zaitsev reports on an InnoDB performance gotcha with larger queries.

Here on the Pythian Blog, Singer Wang unearthed a MySQL 5.1 and InnoDB hot backup gotcha.

Eric Bergen offers his InnoDB deadlock count patch, which he introduces thus: “[Deadlocks] usually aren’t a problem until they start happening too frequently.  . . .  [SHOW ENGINE INNODB STATUS] can be useful for debugging but it’s almost impossible to get the rate at which deadlocks are occurring. [This patch] adds a counter to show table status that tracks the number of deadlocks.”

Baron Schwartz had a script snippet to relative-ize numbers embedded in text to share. Read the rest of this entry . . .

This post might seem outside of our focus, but life brings all kinds of challenges. A friend of mine bought a MacBook when she was on vacation in the USA. For obvious reasons, Macs are more common on the other side of the Atlantic. In Europe it’s still rare to see a person using Mac as a personal computer (no flame intended, just stating a fact).

Her Mac completely broke down. The service guys told her she’d need to replace the motherboard, which would cost almost the same as a new computer. The problem was her Mac wouldn’t even start, and all the data she had on a hard-drive was stuck in the neat white box without any signs of life.

Sure, I said, I’m a computer guy I can recover it, can’t I?

I had never worked with Mac before, so I started with initial research to find out what options I have with hardware available in my home computer den.

I came to know that Mac uses filesystem called HFS+, and it can’t be read from Windows 32bit. Great, I thought, I’ve two options—find someone else with a Mac or get it mounted on Linux.

Fortunately, I have a Linux box at home, so it should be easy. I unscrewed the MacBook, and behind the battery there was 2.5 SATA drive. To be able to connect it, I need the interface between 2.5″ SATA drive and USB. For this purpose I’m using a QCP converter cable, which allows you to connect internal 2.5″/3.5″ ATA/SATA drives directly to USB port. (http://www.youtube.com/watch?v=NMtAPgcMtLQ) I really like this piece of hardware—it’s exactly the kind of gadget you want to have around for saving notebook drives.

After connecting the disk, I found that my OEL5.1 wouldn’t be friends with it. I simply couldn’t find the right hfsplus module for this distribution. Fortunately, there were many references about mounting hfsplus disks on Ubuntu Linux, which is my second system.

I downloaded the required package and dependency libraries for Ubuntu from here:

http://packages.ubuntu.com/dapper/hfsplus

The packages installation is straight forward:

root@silverbox:~# dpkg -i libhfsp0_1.0.4-10ubuntu1_i386.deb libc6_2.3.6-0ubuntu20_i386.deb hfsplus_1.0.4-10ubuntu1_i386.deb

After that, I needed to load the hfsplus module:

root@silverbox:~# modprobe hfsplus
root@silverbox:~# cat /proc/filesystems | grep hfs
        hfsplus

Next, I had to check which partition is the one I need to mount. For this purpose, I used parted:

root@silverbox:~# parted /dev/sdd
GNU Parted 1.7.1
Using /dev/sdd
Welcome to GNU Parted! Type 'help' to view a list of commands.
(parted) print

Disk /dev/sdd: 160GB
Sector size (logical/physical): 512B/512B
Partition Table: gpt

Number  Start   End    Size   File system  Name                  Flags
 1      20.5kB  210MB  210MB  fat32        EFI System Partition  boot
 2      210MB   160GB  160GB  hfs+         Untitled

Knowing the partition containing the data was /dev/sdd2, and I could mount it.

root@silverbox:~# mount -t hfsplus /dev/sdd2 /mnt/macosx
root@silverbox:~#

The next problem I faced was privileges. The directories I needed to save were owned by a non-existent user, and so I wasn’t able to access that path.

To work around this, I created a new user and assigned the directory owner UID.

root@silverbox:~# useradd macuser
root@silverbox:~# usermod -u 501 macuser

This allowed me to access the directory I needed to recover, and copy files to another ntfs disk which will be readable by regular Windows machine.

Update 24-06-2010: Peter de Vaal reported that these instructions work without changes for APEX 4.0 as well. Great news!

Few days ago, I have put together the Quick Install Guide for Oracle 10g Release 2 on Mac OS X Leopard (Intel). I did mention that it would be cool to get APEX working as well but, apparently, APEX isn’t supposed to be running on Oracle 10g Release 2 Database on OS X as few people pointed our on the blogs.

Well, the release notes says that Oracle Application Express is not supported and I would imagine that it’s because Oracle HTTP Server is not supported either. I’m not a frequent user of APEX but I know that APEX 3.2 is not supposed to be configured in Oracle 10g Database using Embedded PL/SQL Gateway. However, looks like few people did manage to run APEX on 10g using Embedded PL/SQL Gateway. For example, Jeff Kemp has some APEX on 10g hints that I perused.

Long story short, I was able to install APEX successfully using the native Mac OS X Oracle 10g Database and here is how…
Read the rest of this entry . . .

I was very surprised that Oracle released Oracle Database for Mac OSX, especially, version 10g now that 11g has been out for almost 2 years. Well, I guess Oracle wanted to please Mac users expecting things just work and decided that good proved 10g is the way to go. On the other hand, we’ve been supporting 11g in production for quite a while and I must say it’s much better quality compare to 10g when it came out. I’m pretty sure there was a significant customer that influenced that decision — interesting who might that be?

Anyway, there is no quick install guide for OS X but only a standard Oracle® Database Installation Guide
10g Release 2 (10.2) for Apple Mac OS X (Intel). It’s fine but if you want to install Oracle on your MacBook and not for production use then you might take some shortcuts and follow a quick instructions so I gathered my notes while installing just released Oracle Database 10.2.0.4 on my MacBook and this is what you see now.
Read the rest of this entry . . .

This topic has been raised again and again and quite a few people have asked me how to configure RAC on VMware Fusions on Mac. This warrants a blog post, especially, that Mac is definitely the way to go for an Oracle DBA — a Unix desktop OS that just works. What can be better? Sorry, I digress without even starting!

Before I go any further, I should say that this is not a complete guide on the Oracle RAC install with VMware Fusion but just the hints on setting up shared storage for Oracle RAC using Mac as host for VMWare Fusion virtual machines (VM’s). The reader is assumed to understand how to setup Oracle RAC and has general understanding of VMware itself. There are plenty of guides on the Internet on how to setup Oracle RAC including VMware but they usually refer to VMware Server on Linux or Windows. Please note that I’m writing it largely by memory so if you hit any issue — please leave a comment.

Disclaimers are over — moving on!

The root of the problem is that VMware Fusion doesn’t support shared disks unlike VMware Server on Windows and Linux. If you try to update the .vmx file manually to enable shared disk, you get the error message “Clustering is not supported for VMware Fusion – this setting will be ignored”. Fear not — you are running the best desktop OS anyway! ;-)
Read the rest of this entry . . .

Previously on Log Buffer: Log Buffer #119.

And now.

Welcome to Log Buffer #120. My name is Warner, and I’m a SQL Server DBA at The Pythian Group. This is my first time on Log Buffer duties ever, so here’s hoping I can give everyone a fair and unbiased look at this week in the database blogging world (and related).

I admit I had no idea of the community or state of the PostgreSQL RDBMS, and so I definitely learned some new stuff this week. First off, over on “The Scale-out Blog” Robert Hodges invites us all to get our shoephone and get smart about the new world of PostgreSQL replication.

Moving over to “ad’s corner”, Andreas Scherbaum gives us a glimpse of the glitz and glam of PGDay opening. Then he entices us with the title, “Party in the evening” just to horrify us by revealing that—you’d better sit down—by the end of the event half of the beer was still untouched. Next time Andreas, who you gonna call?

“Everything is a DNS problem,” is my new voicemail message, and also the title for Kris Buytaert’s blog, where we go deep into uncovering once again the 10 month-old enigma of why did Sun buy MySQL.

Next up, I will speak to you, Mr. (or Ms.) Cross-Platform DBA. You think you know all of Oracle’s exp command-line switches? All of SQL Server’s bcp formats? And all of the things you can do with . . .  whatever you use with DB2? Read the rest of this entry . . .