Every 'optimization' recommendation
would be specific to your data / view(s).
So it is hard to suggest anything
without actually looking at the wb.
Could you please share some?
Besides, the Performance Recording wb
would be of some help (at least).
I'd also like to add to what Yuriy mentioned,
- I'd be curious to know, if you can aggregate your data at database level? ETL would be something to look into. Performance also depends on the calculations Tableau is performing to aggregate data and visualize it.
- You mentioned creating an extract and it didn't help? What server version are you on? I would, without any hesitation say Hyper extract is way faster. I tried and tested Live vs HyperExtract for POS data and the performance difference is significant.
- At the end, it all depends on the end requirements and data structure. So, what worked for me wouldn't work for you.
HI Yuri & Shivaram,
Thanks for your reply, I analysed further the dashboard and I found that about 50-60% of calculated fields can be precalculated. We are using hyper extract. So I guess pre-calculating the fields will reduce a lot of workload rather generating them on-the-fly calculation.
I will keep you guys posted how did go.
I have found the solution for the problem.
The approach of using context filters is correct, I found that there were plenty of LODs (by previous developer) which can be calculated in the Tableau calculations so the response time changed to miliseconds rather than of being a 40 second painful wait per click .