1 of 1 people found this helpful
Heya Dale -
"Refresh less often" means Cached data sticks around for up to 12 hours. There are a fair number of situations where the cache is expired and/or lost..
- The data source is modified
- A viz with an embedded data source is re-published
- Tableau Server is restarted
- An extract is refreshed (this counts as a modified data source as far as we're concerned)
Thanks for Reply Russell. We would like to keep data cache for a week.
Do you think , if i select Balance option and fill 10080 Mins ( 7 * 24 * 60) . The data cache will not get lost for a week, if we control other situations of data lost.
Actually . Our users subscribe to reports and then we update them on weekly basis.
Is the desired result better performance, or greater control over what data the user sees in the view?
If the desired result is performance, you may want to look at:
Additionally, for a better understanding of our caching behavior, this is a good article from a third party:
If the desired result is that you wish greater control over what data the user sees in a view, are you using a live connection to your database, or an extract? If you're using an extract, the data underlying the view will remain persistent until the next time the extract is refreshed. For additional information on how and why to use extracts, please see:
Hope this helps!
Hi Chris. I came across your reply and would like to qualify "the cache is expired when an extract is refreshed". This doesn't seem to occur for our published extract that is refreshed once an hour. Our workbook that points at it has stale data, and it's an embedded view so the toolbar doesn't appear by design. Is the rule different for published extracts vs. embedded extracts? My clients are asking me to add a ?:refresh=yes onto the URL, but I'm gun shy. If the refresh of embedded extracts causes cache to expire, I may go toward this option instead.