You and i – ARE Solving Problems
Do you work in IT? Does this scenario sound familiar?
User: The application stopped working.
You: When?
User: Just now. It was working fine just yesterday.
You: What did you change?
User: Nothing.
You [in your mind]: Yeah, right.
When I talk to customers, they almost universally recognize this situation.
So what do we do when these sorts of problems pop up? Well, we know that applications don’t just stop working. Something changes, and the change causes a problem. The user might have done something, or an administrator may have changed something that affected the user. Or a programmer may have changed the application. Whatever the case, something changed. We have to find out what that is. Determining what changed has gotten increasingly complicated, as new programming environments are combined, as new function gets used, as underlying technologies increase in complexity.
In IBM i development and support, we have the same scenario, so some of our developers got together and created a program that captured some of the application environment when things were running well, and then captured those same attributes when something went wrong for comparison purposes. Our teams who used this significantly decreased the amount of time it took them to diagnose “user errors.” It was so successful that they proposed making a product out of it.
Thus came into being Application Runtime Expert for i (ARE).
When I describe ARE to people, I talk about taking “snapshots” of the system. ARE can take a snapshot of a system (a partition, really) when things are running OK, then take another snapshot later and compare the two, highlighting changes. Since problems tend to be related to changes, ARE helps narrow down problem discovery. All by itself, that makes ARE a great tool, and the kind of integrated ease-of-use tool we IBM i developers like to create.
But ARE can do more than that. You can compare a snapshot of one machine (partition) to a snapshot of another. That’s right – you can compare one machine to another. And, in fact those two machines can even be at different releases. So did you just install 7.1 on one machine and something is not working? Have ARE compare that 7.1 system to a 6.1 machine where the application is working. There are even some things – authority settings – that ARE offers to change back to their original settings for you.
What does ARE include in its snapshots? Many, many things! There’s an initial template with things like File/Directory/Library authorities, System Values, Product Install status and much more (see the product page for a more complete list). These will cover many of the runtime attributes that affect applications.
Even better, though, is that you can customize the snapshot by adding things to the initial template. If you have your own application and know that a particular file, directory or program needs to be set up a certain way, you can add that to the template. In fact, ARE can be used to create an automated knowledge database. You can start very simply with the attributes you include in the template, then as problems occur over time – problems that are caused by the application environment rather than bugs in the application – you can add the detection of those problems to your template.
The ARE interface is browser based, and can run on IBM i 6.1 and 7.1. It can take snapshots (create templates) for those releases, and for 5.4 (i5/OS V5R4) as well. It’s a separate product, so see the product page for more details.
We think this tool has a good chance of making day-to-day life better for people IBM i IT shops everywhere. Take a look. See what you think. And if you have suggestions for additional function, let us know. The team already has a bunch of ideas, but we’ll happily entertain more. We want to ensure that You and i ARE making IT easier.




I believe this is an excellent tool. I could see that triggers can be invoked upon specific difference (from snapshots). Triggers can operate the change or prompt the user. Following this idea a standard installation wizard can be made for any product: ptf, IBM, ISV product or combination.
Posted by: Pascal Polverini | 07/01/2010 at 04:34 PM
Aside from this being a cool product, what interests me more is how it grew organically within your walls. Do the IBM i teams at IBM get to spend a certain percentage of their time pursuing their own creative ideas similar to what Google allows? If anybody reading this doesn't know what I am talking about, you can learn more about it here: http://www.scottberkun.com/blog/2008/thoughts-on-googles-20-time/
It might just be my perception, but Google seems to be doing better at rounding out ideas than any other company as it relates to common business function software. Gmail and GoogleDocs are perfect examples of this IMO. But Google isn't stopping there. They pose a threat to operating system runtime environments with the Google Apps Engine (http://code.google.com/appengine/).
They don't stop at harvesting ideas from the 10k employees but let the public also contribute through their Google Code initiatives: http://code.google.com/. Now, IBM does have something similar titled DeveloperWorks (http://www.ibm.com/developerworks/), but there is a certain OS, DB and language IBM i, DB2 and RPG missing from the left nav on the site. Note that DB2 for i is mentioned (i.e. http://www.ibm.com/developerworks/data/products/db2i5OS/), but not without digging for it.
My question: What are ways that we could get better representation on the DeveloperWorks site so it becomes a breeding grounds for IBMi related software?
AaronBartell.com
Posted by: Aaron Bartell | 07/05/2010 at 07:43 AM