Categories
- All Categories
- Oracle Analytics Learning Hub
- 25 Oracle Analytics Sharing Center
- 17 Oracle Analytics Lounge
- 233 Oracle Analytics News
- 45 Oracle Analytics Videos
- 15.9K Oracle Analytics Forums
- 6.2K Oracle Analytics Idea Labs
- Oracle Analytics User Groups
- 87 Oracle Analytics Trainings
- 15 Oracle Analytics Data Visualizations Challenge
- Find Partners
- For Partners
Oracle Analytics cloud (classic) : Enhance report Excel export timing
Hello,
We have these reports on OAC classic, on which we recently did a global optimization starting from source data to the reporting on OAC.
Loading times went from 2, 3 minutes to seconds for every page so we from our end there is nothing we can do to improve the timings,
The problem is that when we export the report it almost takes 20 minutes which is huge when you know that each page loads instantely.
In conclusion, we have implemented all possible optimizations on our side.
We really need your support to improve export performance — this is not the first time we face this issue.
The export engine itself needs further optimization, including parallelization for better performance.
OAC (both Classic and DV) is an excellent tool, but the slow export process is becoming a serious concern.
Some clients are starting to consider switching to other tools because, despite our best efforts, we cannot improve export speed any further on our end.
An SR has been opened regarding this topic :
4-0001170456
Regards,
Chaima.
Comments
-
I have a suggestion for you despite my better judgement. You could consider using an "agent" (the classic version) and have the jobs scheduled to export and send to your email.
The bigger question is, why are you exporting large amounts of data out of a reporting system? If you can determine why people are doing that, you can more accurately get to a solution. I'm my experience exporting to Excel means they can't get their answers from the report. Then they end up using Excel as their reporting tool instead of the reporting system where the data already lives.
My suggestion is to modernize and get out of classic and into DV. Next steps are to start exploring the AI features available in DV. Modernize first before you try to solve the problem with outdated technology.
0 -
Hi,
The main issue with scheduling is that users no longer have the ability to export reports at any time, which means it’s not guaranteed that they’re working with up-to-date data.
I wouldn’t say that we’re exporting a large amount of data — the Excel file is only about 300 KB — yet it still takes almost 20 minutes to generate. Users continue to export reports because, due to tool limitations, some modifications still need to be done manually. Unfortunately, this limitation also exists in Data Visualization (DV), otherwise we would already be ready to modernize.
Say we have a Contract dimension that includes two sub-levels:
- Contract
- Contract0.1
- Contract0.2
- Contract0.1
Currently, our analysis displays data grouped by the Contract level, followed by all members at the Contract0.1 level, and then all members at the Contract0.2 level.
After exporting, users manually reorder the analysis to align each parent contract with its corresponding child contracts — moving from one parent to its children, then on to the next.This type of hierarchical ordering is not possible in OAC, which is the main reason we cannot fully migrate to Data Visualization.
0 - Contract
-
Are you using the RPD or Web Semantic Model (WSM) for this dataset? You should be able to create that hierarchy within the subject area model.
0 -
Hey Branden,
I’m using an RPD connection between OAC and EPM Cloud.
The hierarchy does exist, but when I display the members in the section compartment of a table, it shows all members from the first level, then all from the second, and so on.What I actually need is for the report to display all descendants of each member before moving to the next parent — essentially following the natural hierarchy order.
0 -
Are the different contract levels all in the same column or are they separated into multiple columns? If you are able to mock up a sample scenario with fake data, that might help me better understand the request.
0
