Taking the guesswork out of upgrade planning

by Pat Neary

When Lyle Ekdahl, SVP and General Manager Oracle JD Edwards, gave his keynote at last year’s Collaborate event, he shared his vision for the JDE E1 roadmap. 9.2 would be the last major upgrade, with future enhancements released under a continuous delivery model.

At the time, he shared some interesting statistics. 9.2 was released in October 2015, by April 2017 around 25% of the JD Edwards installed base had migrated to the latest release. Oracle predicted this would grow to 60% over the coming 12-24 months, most of which would come from customers currently on 9.1.

Our own research suggests the adoption rate has been less than expected. In a survey carried out in January 2018, results indicated there had only been modest growth, with 35% having upgraded to 9.2.

It’s not unusual for JD Edwards users to want to squeeze the last ounce of value from a release. Many customers will stick with a legacy version until it reaches the end of extended support. As 9.0.2 reaches this watershed later this year, we may see a renaissance in upgrade projects in 2018.

Given the number of enhancements released over the past 12 months, it might seem strange that the majority of businesses haven’t chosen to upgrade yet. It’s possible that some may have been waiting to see if Oracle’s commitment to 9.2 and continuous delivery remained unshakeable.

However, it’s more likely that they are wary of change. Specifically, they may want to avoid the perceived cost and complexity involved in a major upgrade. I say perceived, because in many cases, businesses struggle to accurately predict the time and effort required for an upgrade – particularly when it comes to uplifting custom items or extensions.

Just how big is your modified footprint?

Accurate information is the key to good decision making. When assessing the potential impact of an upgrade, you need to clearly identify what, and where, your mods are. Traditional methods, such as CNC reporting, can create inaccurate estimates as they over, or under-estimate the number of modified objects.

False flags can lead to an over-estimation of the number of modified objects, exaggerating the size of your modified footprint by anything up to 60%. Worse still, a missing flag could significantly impact on systems performance if it’s missed during the technical uplift.

The type and extent of modification will also impact on the amount of effort required to uplift the code. Custom code, modified standard code or modified copies of standard code all have varying degrees of complexity.

At DWS we do things a little differently. Our Dimension Analyze™ service enables customers to audit and scope their E1 upgrades, down to an unprecedented level of detail. We analyse every object, line of code and setting, down to pixel movement level of detail.

We identify the from/to base net change for every modified object and we identify the net change type and severity of impact against every modified object. We also identify all those modified objects that are no longer in use, so do not need to be upgraded.

What is the best upgrade approach?

How do you decide on what is the best upgrade approach for your modified code? Do you take the new object and reapply the changes you made? Or do you take your revised objects and apply Oracle changes to your version? How do you decide which will be quicker?

How do you avoid missing a custom object that references a standard JDE object that has become obsolete, or no longer exists in the future release, and becomes an issue during user acceptance testing? What about changes to standard hooks? There may be a change in the parameters passed into a function call, which ultimately affects the results of using this function in custom code.

Knowing about these changes, and the potential impact during the retrofitting of an object, minimises or even eliminates issues during User Acceptance Testing by enabling you to retrofit the object accordingly first time.

Quantifying the cost of an upgrade

Without an accurate picture of your modified footprint, its virtually impossible to put an accurate cost to the upgrade. Sample and extrapolate, or worse still the “one-third rule”, have proven wholly inaccurate when it comes to estimating the time and effort required for a technical uplift.

Dimension Analyze eliminates the guesswork in upgrade planning. With the detailed information it provides, we can estimate the upgrade effort required for every modified object; not just in man days but right down to hours and minutes level of detail. This means we are able to offer fixed price and timescale upgrade services.

Discover the benefits DWS were able to deliver for Chivas Brothers utilizing Dimension Analyze

Find out more about Dimension Analyze or contact us to discuss your requirements.