One of the databases that handle the backend storage of sessions blocked writes which caused a slowdown in availability of session data to the Data API. Failsafes built into Learnosity's system ensured that there was no data loss, and sessions that could not be written immediately remained in a queue for processing. Rebooting an RDS instance solved the issue and the remaining sessions were then quickly processed and the queue was cleared.
Because there was no data loss, this issue temporarily affected availability of session reports/reviews, had a possible impact on adaptive sessions where scoring decisions affected next item delivery and resuming sessions during this period, and may have contributed isolated effects to the assessment stack where custom implementations queried session results before follow-up actions could occur (such as determining which assessment to deliver based on the results of a preliminary diagnostic assessment). We apologize for any inconvenience this may have caused.
There were no log entries showing any failures, and other databases were still writing correctly. We are following up with AWS to see if there is any more information available. Thank you.