Refresh staggering across multiple workspace

[rdnguyen]

Hi All,

Kindly help me understand better of how Refresh stagering which is set at least 30 mins apart vs Dataflow across multiple workspaces that is currently promoted as Dataflow resuability?

Assuming I set up 3 tiers, stagering starting to pull in data at 6 AM, then my business logic Dataflow would do their job around 6:30 AM; then at least at 7 my report could run the refresh. How could I mitigate this time gap, not even count in the chance that Dataflow may failed in between for some odd reasons, please advise?

[rdnguyen]

The current refresh schedule set time slot for every 30 mins, so stagering across workspace would be at least 30 mins - in other words, report would have to be 30 mins -  apart despite the fact that the dataflow series could only take 1 min to complete the refresh.

[collinq · Accepted solution]

Hi @rdnguyen ,

Yes, that is the best solution at the moment.  Perhaps you could search the ideas and see if there is a suggestion to daisy chain dataflows or perhaps suggest an idea and we can all vote for it if you put those links here.

[collinq]

Hi @rdnguyen ,

At this time there is no "when this refresh finishes, start this refresh". There is no way to time two refreshes other than to manually mitigate it.  You can do that by seeing how long the Dataflow takes to refresh on average and then give at least that much time for the next report's refresh.

Also, you could have the second report refresh 2 or 3 times in a row if you want to make sure that you are hitting the latest Dataflow data.  In other words, you could have the Dataflow run at 6:30 and then have the report refresh at 7, 7:30 and 8:00.  That way you know you hit it.  This makes an assumption that you are ONLY hitting the Dataflow and that the report does not have other data sources that might make it take a while to refresh.