So my admin setup a sandbox environment and loaded 10.2 so that I could work on Dynamic Cubes.
I published the cube to the Query Service, and it started correctly, but when I attempted to build a report off of the cube and then run the report...i do not see any performance advantage. This made me take a look at the metrics within the Dynamic Cube from the Query Service in Administration. Here is where I noticed that the Data Cache Hit rate,the In-memory aggregate cache hit rate, and the Result Cache hit rate were all at 0%...this was after I attempted to run the report a few times.
I am not sure if this is normal! If anyone has experience in configuring the Dynamic Cubes and Query Service...please let me know.
THanks!
You will need to load your in memory aggregates to notice a difference. Othwerwise it will always make that round trip to the db
Thank you for your quick reply!!
So how is the in memory aggregates loaded? My Dynamic Cube is a very simple one, in fact it is a POC to demonstrate to the business the speed it should provide. I have only one measure, but I have 2 very large dimensions where the measure needs to aggregate against...
Please advise if this loading is a step that is included when publishing the cube...basically when I build the cube within Cube Designer...I was under the impression that the data and aggregate cache was being created???
You will need to download and run the DQA tool. The in-memory aggregate definitions are loaded into the content store from this tool.
Thanks!
So without loading the aggregate cache, I will not see any significant changes to the performance? I just want to get clarity on this subject so as to better explain to my business!
As I mentioned, the metrics also show 0.0% Data Cache hit rate as well...is that normal?