nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Oracle Communities

Oracle Analytics and AI

Child Item

Option for Data Viz data sets to be in-memory columnar storage

Antony Heljula

Organization Name

HSBC Plc

Description

Hi, we received feedback from our users that the responsiveness/performance of Data Viz was seemingly quite slow in relation to their Qlik dashboards. We therefore asked for a demonstration of their Qlik dashboards which were based on an underlying data set of 3.5M rows with about 40-50 metrics/attributes.

The performance was lightening fast via their web-browser. When a user drills or selects a filter the refresh is so fast it is easy to miss it, even complex filters and time series functions responded instantly - literally in the blink of an eye.

They pointed out that the performance is mainly due to the way Qlik stores its data set as an in-memory columnar data store.

So our idea is for Oracle Data Viz to provide the option of delivering its data sets as an in-memory columnar data store, to optimise the existing row-based data set performance.

Thanks!

Use Case and Business Need

Additional capabilities to compete against competitors, improved user engagement with lightening fast response times.

Original Idea Number: 51fdfba0ec

Find more posts tagged with

Delivered

Reporting

Data Modelling and Datasets

Analysis and Data Visualization

Developers

Status: Delivered

Comments

Viv

Data refresh are analysis is slow at times and would be good to look at how this can be improved.

M_Kumaran

I believe in-memory columnar storage will enhance the performance of the visualizations.

Angel Shipp

I agree with your sentiments. Do you have insider knowledge on when this planned capability is slated to be released?

Angel Shipp

Gabby Rubin, do you have updates on when we can expect to have this update?

vkm

This has been a real pain point for our users

Gabby Rubin-Oracle

The new high-performance caching engine is being rolled out and is in most regions at the moment. Note that given that the engine is part of the OAC infrastructure, there is no visible difference, but hopefully, you will see the performance improvements. As always, mileage will vary and every data management technology has areas where it is better or worst. If you experience degradation, please let us know and open an SR.

Michal Zima

Hi Gabby,

you are mentioning, that: "engine is part of the OAC infrastructure" . From this I am deducting, that this "feature" will be available solely in OAC . Or is there a plan to include this engine (as "option") also in Oracle Analytics Server (on-premise sw) ?

Thanks in advance for you answer and kind regards

Michal

Angel Shipp

Hi Gabby Rubin, do you have documentation to share about the 'High performance caching engine"? We are interested to know the details behind it so as to not inadvertently submit ideas that are already part of this caching engine.

Rajakumar Burra

Performance is still an issue if we compare with Qlik.

Qlik filtering is much faster than OAC. If we have more calculations , OAC is much slower than Qlik.

Ambient

@Rajakumar Burra

I just stumbled onto this thread...

What kind of caching strategies were you using to increase the performance?

How do OAC cache config strategies compare to Qlik's, in terms of complexity to implement?

https://docs.oracle.com/en/cloud/paas/analytics-cloud/acabi/manage-query-caching.html#GUID-C9C7EBEE-4C49-4F7F-980E-92A1DCFE24A3

https://docs.oracle.com/en/cloud/paas/analytics-cloud/acabi/configure-advanced-options.html#GUID-6AD28E80-87E9-42FB-9B4A-0C72149EDF98