IBMDatabaseMagazine.com Blogs

Last month we took a look at Data Partitioning, an Informix Dynamic Server feature that gives you the ability to break your larger tables into seperate partitions. We saw how this feature can be used to quickly detach and drop older data from your tables and attach new partitions to store new data giving you a convenient way to always keep the most recent 6 months of data online.

What isn't so convenient about using a partitioning strategy to accomplish this goal is the management overhead (the DBA's management overhead, not the engine's) required to drop and add these partitions on a daily or even weekly basis. Being an Informix DBA I want to continue to live and work in the "Administration Free Zone" so I created a python script to do this work for me. Now where is my 5 iron...

Continue reading "Scripted Detach/Attach Partition with Informix Dynamic Server..."

Comment on this blog entry


Data Partitioning with Informix Dynamic Server

One of the engines that I manage contains a particularily nasty table. Let me know if this sounds familiar: This troublemaker is a history table that stores the details of each of the 2 million transactions our platform processes each day. The rows are wide, weighing in at over 900 bytes per row, letting me fit a whopping 2 rows per 2K page. The table contains the most recent six months of history; anything older than six months must be removed nightly to make room for tomorrow's data.

Without data partitioning this table could not exist.

Continue reading "Data Partitioning with Informix Dynamic Server..."

Comments(4)


Informix security frameworks: Just how difficult is it to retro-fit "security"?

Everyone "knows" it's much easier to design and build a system that is secure, and that it's much more difficult to retrofit security to an existing environment. But how difficult is it really?

Continue reading "Informix security frameworks: Just how difficult is it to retro-fit "security"?..."

Comment on this blog entry


The Informix High Performance Loader - Part 3

Welcome to the third and final installment of the Informix HPL blog series (at least for now.) In parts 1 and 2 we abandoned the GUI and went old school command line with onpladm (that's the way it was and weeeeeeeee liked it!), we unloaded and loaded lots of data to and from a flat file on disk, we moved data from one table to another in parallel using named pipes and we even moved data between servers without using flat files. This toolset should be enough to cover most of what you would ever need to do with HPL. If not, it will at least give you a solid foundataion to start with. In my final attempt to increase HPL awareness I want to take a look at the more tedious, but sometimes necessary, HPL maps and formats objects.

Continue reading "The Informix High Performance Loader - Part 3..."

Comment on this blog entry


The Informix High Performance Loader - Part 2

Last month we explored some of the basic, but powerful, funtionality of the Informix High Performance Loader (HPL) and the CLI configuration utility onpladm. Today we will be looking at some advanced uses of HPL that will allow you to simultaneously unload and load data without the need for large intermediate data files.

Continue reading "The Informix High Performance Loader - Part 2..."

Comments(2)


Informix basks in the iPhone's glow

With Apple's latest announcements making headlines (hear about the cheaper iPhone, anyone?), it's worth mentioning Informix Dynamic Server's Mac OS X support.

Continue reading "Informix basks in the iPhone's glow..."

Comment on this blog entry


Celebrate DB2, U2, IMS, and IDS anniversaries by helping write their histories

During this week 25 years ago (June 7, specifically) DB2 made its debut (check out the original press release for a trip down memory lane) on the mainframe. As storied as its long history is, DB2 isn't the only IBM database technology celebrating a milestone this year. It's joined by UniData and IMS (20 and 40 years, respectively). Of course, although they're not celebrating anniversaries that end in 5 or 0 this year, Informix (27 years) and UniVerse (23 years), certainly have eventful histories.

Now that the IBM Database Magazine wiki is back from a brief cloud-induced hiatus (so long, StikiPad), you're invited to help document the achievements of your favorite IBM database technologies in its pages.

Continue reading "Celebrate DB2, U2, IMS, and IDS anniversaries by helping write their histories..."

Comment on this blog entry


The Informix High Performance Loader – Part 1

The High Performance Loader is the absolute best utility available when it comes to inserting or unloading large amounts of data from your systems. Up until version 9.4 you had two options if you wanted to configure an HPL load or unload job. You could use the ipload GUI or you could reverse engineer the steps required to correctly populate the onpload database.

Continue reading "The Informix High Performance Loader – Part 1..."

Comments(1)


Finding the Right Page Size for Your Indexes on Informix

In my last blog we took a look at storing indexes in a dbspace with a large page size to try to decrease the number of levels in an index and increase performance by reducing the number of I/Os required to traverse the index. Using the largest page size available, 16K, we were able to decrease the number of levels in our test index from 5 to 4, but only realized a slight performance gain if--and only if--we increased our resident segment (buffers) from 512 MB to 1.5 GB to compensate for the larger, less selective page size and larger cache miss penalty of a 16K disk read vs a 2K disk read. I ended the blog with the suggestion that the best solution would be to find the smallest page size that still reduces the number of levels, giving us the benefits of better caching through smaller index pages and the benefits of reduced index levels when using larger page sizes.

Continue reading "Finding the Right Page Size for Your Indexes on Informix..."

Comments(2)


IIUG Power Conference Update: Winds of Change

The first International Informix User Group conference since the IBM acquisition in 2001 kicked off this morning with a keynote by Dr. Ambuj Goyal, general manager of IBM Information Management and self-described doting parent to many RDBMSs, each with its own successful market.

Continue reading "IIUG Power Conference Update: Winds of Change..."

Comment on this blog entry


A Web 2.0 Expo debate: creativity vs. control in the enterprise

Forrester Research is projecting that companies will spend nearly $5 billion on Web 2.0 technologies like mashups, wikis, widgets, and so on in the next five years. That explains the tremendous energy around the Web 2.0 Expo underway in San Francisco this week.

Tomorrow Anant Jhingran, CTO of Information Management, and RedMonk analyst Stephen O'Grady face off over the challenges of bringing Web 2.0 technologies into the enterprise in a session called "Creativity vs. Control: The Debate Continues—Can Enterprises Have Both?"

Continue reading "A Web 2.0 Expo debate: creativity vs. control in the enterprise..."

Comment on this blog entry


Configurable Page Size (CPS) and Informix

In version 10.0 of Informix Dynamic Server we were given the ability configure the page size of the data stored in a dbspace. For my Linux-based servers this means I have the option to store data and index pages on 2K, 4K, 6K, 8K, 10K, 12K, 14K and 16K pages instead of the previous option of just 2K pages. That's all fine and dandy, but how can we put this to good use?

Continue reading "Configurable Page Size (CPS) and Informix..."

Comment on this blog entry


Writing a Simple Informix Engine Benchmark in Python

The kind people at IBM Database Magazine have asked me to join their highly unpaid blogging staff to write a few paragraphs a month on all things Informix. I'll be pulling from my own day to day experiences as a DBA who frequently puts on the software developer, system administrator, database architect and storage administrator hats for content. Today I'll be wearing the software developer and DBA hats to write a quick and easy python script to hammer our database engine to get a rough idea of what kind of load our servers can handle.

Continue reading "Writing a Simple Informix Engine Benchmark in Python..."

Comments(3)


Most read 2007 articles: DB2 and Informix

As we finish out one year and ramp up for another, I took a peek through the Web stats to see what was on visitors' minds when they stopped by the DB2mag site. I found a few surprises. Some of the most-read articles are from years ago. Two focus on Informix (evidence of that our expanded coverage there is finding an audience). Check out the top ten.

Continue reading "Most read 2007 articles: DB2 and Informix..."

Comment on this blog entry


Informix Updates: New wiki, conference registration open, upcoming Webinar

Informix enthusiast Eric Herber has added a wiki to his Informix Zone Web site. The wiki's purpose is to:

provide useful information around IBM's brand new INFORMIX offering. The Wiki covers areas like installing, configuring and developing with the free IDS Developer Edition.

Meanwhile, the International Informix User Group is offering the free IDS 11 Developer Edition download in its members' area. While on the site, you might want to register for the Informix conference next April. There's been a lot of Informix activity this year--who better to learn new features from than Informix developers and gurus. You can learn more about the conference lineup here.

If you can't wait until April to pick up new Informix skills, you do have options. Yesterday, IBM's Grant Hutchison (the IBM Data Studio product manager) posted these tips on using Data Studio with IDS 11. Grant will join Guy Bowerman to present a webinar on "Fastrack for Application Development" next week. Registration is open for that event.

Comment on this blog entry


RFID

RFID is everywhere at IOD. I love RFID, I think this is a great technology. I really wish to find and work on an RFID project (call!). Of course, there are some privacy issues, but I think RFID will grow as much as cell phones in the 90s.

Continue reading "RFID..."

Comment on this blog entry


CAC

CAC stands for Customer Advisroy Council. It is a group of people who wants to participate in a closer way to the development of IDS.

Continue reading "CAC..."

Comment on this blog entry


Buffer Pool

Some are concerned by such things as buffer pools. I am not.

Continue reading "Buffer Pool..."

Comment on this blog entry


Heading to LV

I am on my way to the IOD conference. This year it is in Las Vegas. I just can’t wait to see Céline Dion on stage.

Continue reading "Heading to LV..."

Comment on this blog entry


Informix User Head Count: Beyond the Numbers

A recent story in Computer Business Review online called
IBM has lost 80% of Informix users in six years claims that the number of Informix users today is around 20,000, which it says is down from 100,000 at the time of IBM's Informix acquisition. OK, before I start, let me ask you...what is a user?

Continue reading "Informix User Head Count: Beyond the Numbers..."

Comments(1)


Cheetah is now out of the gate

Well June 12, 2007 Informix Dynamic Server 11 - Cheetah is now back in the forefront and fully announced.

Great features have been added to an already great product. This release of IDS is the best I have seen since the days of the release of IDS 7. IBM has put its money where its mouth is as they saying goes, or in the case of IBM, they put out their best technology foot forward.

Informix Dynamic Server now has features not available in any other database product and now the product with the tag line of Set it and Forget it gets even better. When we all go home from work each night we definitely dont have to be concerned with whta is going on back at work when it comes to the database.

This release of IDS is not only the most packed with features, but the most hyped release I have seen in almost 25 years around the Informix product line.

Special thanks go to Jerry Keesee, Bernie Spang, Bruce Weed, Cindy Fung, Cathy Elliot, Gary Procter, but most of all, and the most important and biggest thanks go to each and every IBM Informix developer. You are the unsung heroes to the Informix community and the heartbeat of this community. I just wish I could list all of you here to thank you.

for more coverage and release details visit the IIUG web site

Stuart Litel
President
International Informix Users Group
www.iiug.org

Run Cheetah Run....

Comments(1)


So tired I could be trashed

I am tired. I am really tired but I just wanted to drop a quick note here, about this great day.

Continue reading "So tired I could be trashed..."

Comment on this blog entry


iiug.tv is born!

Have you been at the IIUG/IDUG Conference?

No, well, sorry, I am getting late putting all my notes together, but that was 'cause of a secret project.

Yes, you were there? Have you seen the funny looking French man, dressed with IIUG cloth (allllll the time) carrying his camcorder around? Well, this was 'cause of a secret project.

Continue reading "iiug.tv is born!..."

Comment on this blog entry


Have you met Stuart?

If you have never met Stuart, because you live in France, or nearby… IIUG and VMark have organized an Informix Day in Paris on May 23rd.

Continue reading "Have you met Stuart?..."

Comment on this blog entry


Back from IIUG/IDUG NA...

Well, as you know, IT projects do not go always right. It’s the same for Blogs.

Continue reading "Back from IIUG/IDUG NA......"

Comment on this blog entry