The Cognito User Sync job has been changed so that it runs immediately in Dev, Test and Prod environments. However, any environment beyond that will have a 10 minute delay as part of the job before completing.
What was the issue?
Cognito in AWS has a limited amount of API calls that it can handle over a set period of time. If the lower environments all run at the same time as Prod and each other, it can cause the job to fail as the limit can be exceeded. The jobs often run at the same time as each refresh sets the lower environment to have the same job schedule as Production.
Why did we delay the start of the job?
As a fix, we have delayed any environment beyond Dev, Test and Prod to 10 minutes after the job is started. This will make sure that limit is not exceeded and that the job does not fail due to exceeding the API Call limit.
How does this affect you?
If you have non Prod environments beyond Dev and Test, the Cognito User Sync job will have a 10 minute delay. This can give the appearance that the job is taking longer when it is actually by design.