Skip to main content

2020 Vision: Optimize System Configuration Settings

In this i Can blog, IBM i expert Dawn May discusses why you shouldn't keep your system at default values.

i Can on a purple bar against a geometric background.

I don’t have 20/20 vision. In fact, I’m very nearsighted and depend upon my glasses to see clearly. When I have my annual eye exam, a number of dials on the phoropter are adjusted, and based upon my feedback, the eye doctor determines the necessary prescription for my eyeglasses.

System configuration settings, like an eye glass prescription, are parameters that must be tuned to your workload and require review on an ongoing basis.

IBM does a pretty good job at the shipped defaults as these defaults are generally considered ‘safe’; the system will work fine without any changes. However, these default settings are often not optimal, particularly as environments scale. More processors, more memory, more disk, more users, and more jobs require important changes to many system settings from their IBM-supplied default.

I’ve worked with many clients since leaving IBM, and one issue I have repeatedly encountered is system configuration parameters that remain at their default setting.  Continuing from my analogy above, this is the person who has never had an eye exam and doesn’t realize their vision is getting fuzzy even as they strain to see. Suboptimal system configuration settings often result in the system doing a lot of unnecessary work.

There are many configuration settings where IBM has documented best practices and recommendations and below are several system configuration parameters you should review to ensure you have set them appropriately for your environment.

IBM documents several system value recommendations in Best Practices for Managing IBM i Jobs and Spooled Files (while seemingly dated, this Redpaper remains relevant).

I often encounter prestart job settings left at the system default settings. You should tune your prestart job entries based upon your workload. IBM documents their recommendations in the Knowledge Center article Prestart Job Tuning and the Tuning Prestart Job Entries experience report.

The following previously published blog articles also review configuration settings:

·      Maximum Spooled Files by Job

·      Job Log Pending

·      Job Log Output

·      Prestart Job Configuration Settings

·      Maximum CPU and Maximum Temporary Storage

·      Why You Should be Using Expert Cache

·      Detach Spooled Files From a Job

·      Avoid Locking Libraries in the Library List

·      The Secret History Log Enhancements


One final note: With IBM canceling/deleting/removing the i developerWorks site, you can find the ‘i Can Blog of Blogs’ at its new home: https://dawnmayi.com/i-can-blog-of-blogs/.


IBM Systems Webinar Icon

View upcoming and on-demand (IBM Z, IBM i, AIX, Power Systems) webinars.
Register now →