If you have admin rights, you should be able to see the execution of the subscriptions within the backgrounder non-extract dashboard. I just looked at mine, the schedule is defined as parallel, and it does appear that there are some subscriptions running in parallel, but we only have 4 backgrounders.
1 of 1 people found this helpful
We have 4 background tasks and lots of schedules running at various times of both subscriptions and extract refreshes which are set to 'parrallel'. When it comes to extracts, I have definitely observered a the tasks within a single schedule set to parallel using all 4 of the background tasks.
When it comes to subscriptions though, it's a bit harder to tell from what I can see of our data (because the subscriptions run at a similar time to extracts and are of lower priority). Looking closely I can see that the subs are running on different backgrounders at times which suggests there may be some parrallelism happening as Jeff has seen. However, if there are mutiple users subscribed to same report on the same subscription schedule, these all seem to be using the same backgrounder, which seems to support your observations too. Might be worth raising a ticket to Tableau Support to get the definitive answer and then post back here so we can all know.
Tableau doesn't support cache pooling, it is user based, that means the same report will be cashed once for each user instead of caching it once and use that cache for all users. I have had the same problem with subscription as well. One of our report had 90 subscriptions and it would take about an hour and half for all emails to be sent out. Therefore I created a separate user (service account user) and created subscription for this user. next I redirected the new subscription from this new user (service account user) to all 90 users. It is a work around so Tableau server will only sends one email subscription instead of 90 emails.