【The definition of improvement】

We can't found anything common characteristics in terms of "design" and "functions used" for OAC reports that increased BI (OAC) 's response time compared to BIEE.

The only commonality was that those reports has a large total number of column x row cells.

【Enhancement Request】

We would like to see doing enhancement made so that OAC extraction performance is achieved even for reports with a large number of rows and columns.

 【Duplicability】

We extracted the data several time, but OAC response time was not changed.

 【Request detail】

We replace the most important analysis system to Oracle Cloud from On premises.

We selected to use Oracle Analysis Cloud as a BI tool.

The system required high performance by users.

So we set the goal as an equal with previous system response time.

And we decided tolerance below.

[Within 10% from the goal]

[Within 2 seconds from the goal]

[Within 5 seconds the time of response]

But ,about some partial reports, we can’t reach the goal or tolerance.

And we raise the oCPU from 4 to 16, but the response time was not improved.

We requested to analyze the reason to Oracle SR member.

The finally, we got the analysis below.

We can't see any different behaver from the specifications as a product.

We discussed what we could do to isolate bottlenecks and develop workarounds at the field level.

Oracle Consulting conducted an initial investigation with application developer to see if there were any hints of improvement that could be gleaned from extracting sample reports and analyze the report definitions.

The results, the only commonality was that those reports has a large total number of column x row cells.

Our Management member and Oracle Japan management member discussed how to solve this issue.

And we decided to raise this enhancement request.

Please solve this performance issue.

【Example】

For example some of the operation that doesn’t reach the goal are used as a below business needs.

Consumer Purchase analysis

->we export the data to excel, than we combined another data and create feedback data for each branch offices.

We use this information to make decisions at our sales offices and management departments, and to provide feedback to our members.

This analysis has being operated about 40,000 times by 630 users as monthly, so there is stress and decreasing productivity due to performance deterioration.

For example, some reports that used to run within 1 minute now take several minutes or even more than 10 minutes.

     If there is performance deterioration about 2 minutes as an average of all , I can calculate that our productivity is already decreasing 80,000 minutes (1,333 hours).

Confirming fixed plan

->we confirm whether input fixed plan data are correct or not in this table.

We use this information for operations that affect sales management plans.

It used to be possible to re-plan many times with a processing time of about 1 minute, but now it takes more than 5 minutes to process.

As a result, the number of runs is limited, impacting lost business opportunities.

Supply Planning Analysis

->we analyze plan and result of production, sales , inventory than, fix the plan.

This analysis impact for our business.

Originally it took about 1000 seconds to export, but with OAC it takes about 1600 seconds.

 This analysis has being operated about 30 times as monthly, so there is stress and decreasing productivity due to performance deterioration.

 There is performance deterioration about 300 minutes by monthly, so our productivity is decreasing.