Forum Discussion
Optimizing and fast refresh of data connected from athena
- 3 years ago
Sorry, I just updated my previous answer around this.
If you want the fastest reponse time in a report, you should use IMPORT mode to bring your data into Power Query, rather than using DIRECT QUERY.
If your ODBC connector is operating in a Direct Query way, then you'll have to wait for communication and processing with the server every time you update your visuals (slicers, interactive filtering etc.).
Pete
Ok, so the example you have there looks fine i.e. the DAX query is only 20ms, it's the 'Other' part that's making it feel slow for you.
This 'Other' time is usually due to queueing, so check the DAX Query times on all of the visuals and, if there's any that are significantly higher than the others, or over about 150-200ms, then look to optimise the measure(s) in those visuals.
If you're running the ODBC connector in Direct Query mode, then these times will include the communication with, and materialisation of, views etc. If you use import mode this will not be included.
You can also reduce the number of visuals on your report page. The more that have to load, the more queueing is required.
Pete
ok, I'll look into those measures, also If there;s any solution at source as well?
would storing the query results to s3 and then connecting to powerbi a better way?
any heads up on that?
- BA_Pete3 years ago
Super User
Sorry, I just updated my previous answer around this.
If you want the fastest reponse time in a report, you should use IMPORT mode to bring your data into Power Query, rather than using DIRECT QUERY.
If your ODBC connector is operating in a Direct Query way, then you'll have to wait for communication and processing with the server every time you update your visuals (slicers, interactive filtering etc.).
Pete