My dataset runs for 9 hours, generating 1.5M records. now the scheduled refreshed faila. Is there a way to check the cause of the issue? The database is mysql RDS
Hi @ebernardinojr if you click on the ‘Failed’ status it should show you more details about the error it hit. Can you share what it says in there?
Ingestion Id
e35040fa-2d78-4e27-b281-cc6a01a76fa1
Error type:
QUERY_TIMEOUT Learn more
The query to this data source timed out waiting for a response. Check your data source and try again.
Error details:
PerformExtract timed out. Will not retr
this is the complete message, my engineer share this looks like the RDS is not capable of the query
Hi @ebernardinojr - Thanks for posting the question. Can you please give more details on the data sets.
- Are you using any custom sql in data sets.
- When you are refreshing the data set, the sql should be executed in the RDS server. Please login to RDS mysql with admin ( via any client tool like mysql work bench) and capture the sql. We need to understand the sql performance . You can check the explain plan of the sql and understand the performance from DB level directly.
- Check the underline tables which is used in data set, see whether you can create any index or not and also check the statistics of the table is correct or not.
9 hours is concerning time for any data set refresh and there is some issue in the db side not QuickSight.
Regards - Sanjeeb
Hi Sanjeeb,
good day
yes, we are using custom sql.
as it turns out the issue is with the IOPS, we had to increase the capacity of the IOPS from 200 to 400. The dataset refresh that are failing and usually takes 9 hours is now down to less than 2 hours.
we are still working on optimizing the query.
thank you,
efren
1 Like
Hi @ebernardinojr - Thank you for the details. Can you also check the explain plan of the sql to see the cost of the plan. Possibly you can try to built indexes for faster performance as well.
Please mark your suggestion as solution so that it can help wider QuickSight community.
Regards - Sanjeeb