Refresh time differences between scheduled - on demand

[mericok]

Hi,

I have a little bit problem with scheduled refreshing time. I scheduled refresh time on my report but sometimes I need to refresh manually. When I did refresh manually which is "on demand" refreshing finish like 4-5 minutes but when starting the schedule refresh it takes a long time. 15-20 minutes. What can be related the main reason for differences? I share a screenshot of my issue.

Thanks.

[Greg_Deckler · Accepted solution]

That's going to be a backend question for Microsoft to answer. I would anticipate the answer goes something along the lines of, there are two different processes for refreshing data. One, when a user clicks refresh, they know that the user is there waiting so it probably goes to the top of the queue stack to get processed. Versus, in a scheduled refresh, it's probably place lower in the queue because it'll happen when it happens. So, my theory is that yes the refresh process starts but it is probably waiting in a queue for a long time and then it refreshes. So the total time is longer. The other explanation might be the amount of resources dedicated to one process versus the other. Again, background refreshes might be placed in a pool that has less resources because they are not time sensitive like user initiated refreshes. This is just me musing.

[Jassie]

Hi Greg, 

It happened to me with the file that has incremental refreshing applied. Now I cannot be sure if it works properly. Unfortunately, I work with a customer user that doesn't have Premium and I cannot see whether partitions are created or not in SSAS, and main database is in BC SaaS so no profiler. From Power BI diagnostic, it GETs date from Range and receives it in response so I guess all work fine, or? Do you have some suggestions on what else I can check? Thanks