Has anyone worked on scheduling alerts on a dashboard with direct query datasets? This page (Working with threshold alerts in Amazon QuickSight - Amazon QuickSight) says alert rules are checked at a random time for direct query datasets. How is this beneficial in case we are trying to track metrics near real time?
Hello @Priyanka_Bisht, I understand that language is incredibly vague but I do believe it would still be helpful even in the case of direct query. With SPICE datasets it is easier since it can check every time the dataset is refreshed. It is able to handle the functionality due to the fact the dataset is already linked with a schedule. I do think it would still be beneficial to utilize alerts with a direct query dataset, especially if you set the alert to notify you as frequently as possible, rather than once a day.
Another option would be to implement a single dataset that runs refreshes as often as possible and connecting the alerts to visuals built from that. Maybe this would provide more timely alerts for your use case depending on the level of importance for the metric.
Thank you, @DylanM !
Let’s say a KPI (with Direct Query connection) reaches the threshold at 2PM CT and I’ve set an alarm for as frequently as possible. When will I get the alert?
When you say this “Another option would be to implement a single dataset that runs refreshes as often as possible and connecting the alerts to visuals built from that.” Are you suggesting a SPICE dataset or Direct Query?
Hello @Priyanka_Bisht, I’ll test the direct query alerts on my end as well, after reading the documentation further, I am not sure if it will send you an alert at 2pm or not. Let me know if you discover anything as well after implementing it on your end. As for my 2nd suggestion, I was recommending to set up a dataset in SPICE that runs refreshes often. Maybe set to 15 to 30 minute refreshes. Then you could build alerts off of that data if the direct query is not providing the results you need.