Peak Render explained + estimate?

[Anonymous]

Hi,

I've read some posts but I still don't understand.

A) Is a report "rendered anew": at least (!) everytime a user loads the report for the first time ( = not from cache)?

B) do filters and sorting contribute to the rendering, or is the report rendered completely if a unique site visitor visits the site for the first time?

C) how long is the result (again: what does it contain?) "normally" stored in the cache?

D) does each Slice&Dice - or each API-request result in a new render?

E) "how fast" is an A-SKU? "On the average": is it "enough" to handle 10, 50, 100, 500 or 1000 unique visitors to a site during an hour? When do delays above e.g. 2 sec. occur (...only approximately & guessed)

Besides not knowing, I'd be very happy with an answer like this: "When you have max. 60 unique site visitors distributed evenly over one hour, and each one slices and dices on your report for approx 5 minutes, you will not have significant delays showing a single-page-report on an A-SKU".

Bye

Michael

[Anonymous]

Hi Ibendlin,

thank you for answering!

Here's the information to the best of my knowledge:

- I don't know about the eviction risk. I just Googled (ok: I should have used Bing) it - I'm not that far into the Process yet. 

- I'll import the data. Within the first two years: Max. about 1 Mio. Rows. "Tiny" - we use PowerBI for its visuals, not for its analytics-power

- Refresh on the organizations side: 3-4 times a year. On the customers side: A user will use "many slicers" to filter the report. Average time spent with the report will be about 5 minutes.

- Not Premium, but embedding for customers. One publisher with Pro-License; "visiting users" from outside the organization without licenses

Bye

Michael

[lbendlin · Accepted solution]

"embedding for customers. One publisher with Pro-License; "visiting users" from outside the organization without licenses"  - not so sure about that, but I'm no expert in embedding. May want to have that checked. Definitely can't use A SKUs - need E SKUs.

Your requirements sound more like rounding errors to a Power BI environment.

[lbendlin]

Needs more details.  How big are the reports (ie what is the dataset eviction risk)?  Import mode or Direct Query?  How many dataset refreshes, how long do they take? Gen1 or Gen2?

[Anonymous]

Hi Ibendlin,

thank you for answering!

Here's the information to the best of my knowledge:

- I don't know about the eviction risk. I just Googled (ok: I should have used Bing) it - I'm not that far into the Process yet. 

- I'll import the data. Within the first two years: Max. about 1 Mio. Rows. "Tiny" - we use PowerBI for its visuals, not for its analytics-power

- Refresh on the organizations side: 3-4 times a year. On the customers side: A user will use "many slicers" to filter the report. Average time spent with the report will be about 5 minutes.

- Not Premium, but embedding for customers. One publisher with Pro-License; "visiting users" from outside the organization without licenses

Bye

Michael