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

Gabby Rubin-Oracle

@Michal Zima Both OAC and OAS have an internal engine that allows users to cache their dataset entities. While the specific engine is different between the two, the ability exists for both. Marking the dataset entity as cached, snapshot the data into the internal OA engine. As announced in the past, the engine used in OAC is based on joint development with the ADW team, which is why it differs from the engine provided with OAS. Both offer great performance; the key difference is that the one we use in OAC is designed as a microservice and uses the same high-performance HW used by ADW.

@Angel Shipp Since this is not a customer-facing engine (outside of the fact that you can mark dataset entities as cached), we do not plan to document it.

@Rajakumar Burra are you comparing the Qlik cached dataset to an OAC cached dataset or the Qlik cached dataset to the OAC RPD/OAC live dataset? We worked with several customers who moved from Qlik, and in most cases, the issue was that the customer moved from a cached dataset in Qlik to a live OAC RPD, which is apples to oranges. I'm sure that there might be scenarios where there will be cached-to-cached performance differences, and we would be interested in analyzing those.